open-consul/agent/acl_endpoint_test.go

597 lines
19 KiB
Go
Raw Normal View History

2014-08-06 17:30:47 +00:00
package agent
import (
"bytes"
"encoding/json"
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
2018-10-19 16:04:07 +00:00
"io"
2014-08-06 17:30:47 +00:00
"net/http"
"net/http/httptest"
"testing"
2015-05-06 02:25:10 +00:00
"github.com/hashicorp/consul/agent/structs"
"github.com/hashicorp/consul/testrpc"
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
2018-10-19 16:04:07 +00:00
"github.com/stretchr/testify/require"
2014-08-06 17:30:47 +00:00
)
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
2018-10-19 16:04:07 +00:00
// NOTE: The tests contained herein are designed to test the HTTP API
// They are not intended to thoroughly test the backing RPC
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
2018-10-19 16:04:07 +00:00
// functionality as that will be done with other tests.
func TestACL_Disabled_Response(t *testing.T) {
t.Parallel()
a := NewTestAgent(t, t.Name(), "")
defer a.Shutdown()
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
2018-10-19 16:04:07 +00:00
type testCase struct {
name string
fn func(resp http.ResponseWriter, req *http.Request) (interface{}, error)
}
tests := []testCase{
{"ACLBootstrap", a.srv.ACLBootstrap},
{"ACLReplicationStatus", a.srv.ACLReplicationStatus},
{"AgentToken", a.srv.AgentToken}, // See TestAgent_Token
{"ACLRulesTranslate", a.srv.ACLRulesTranslate},
{"ACLRulesTranslateLegacyToken", a.srv.ACLRulesTranslateLegacyToken},
{"ACLPolicyList", a.srv.ACLPolicyList},
{"ACLPolicyCRUD", a.srv.ACLPolicyCRUD},
{"ACLPolicyCreate", a.srv.ACLPolicyCreate},
{"ACLTokenList", a.srv.ACLTokenList},
{"ACLTokenCreate", a.srv.ACLTokenCreate},
{"ACLTokenSelf", a.srv.ACLTokenSelf},
{"ACLTokenCRUD", a.srv.ACLTokenCRUD},
}
testrpc.WaitForLeader(t, a.RPC, "dc1")
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
2018-10-19 16:04:07 +00:00
for _, tt := range tests {
t.Run(tt.name, func(t *testing.T) {
req, _ := http.NewRequest("PUT", "/should/not/care", nil)
resp := httptest.NewRecorder()
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
2018-10-19 16:04:07 +00:00
obj, err := tt.fn(resp, req)
require.NoError(t, err)
require.Nil(t, obj)
require.Equal(t, http.StatusUnauthorized, resp.Code)
require.Contains(t, resp.Body.String(), "ACL support disabled")
})
}
}
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
2018-10-19 16:04:07 +00:00
func jsonBody(v interface{}) io.Reader {
2014-08-06 17:30:47 +00:00
body := bytes.NewBuffer(nil)
enc := json.NewEncoder(body)
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
2018-10-19 16:04:07 +00:00
enc.Encode(v)
return body
2014-08-06 17:30:47 +00:00
}
func TestACL_Bootstrap(t *testing.T) {
t.Parallel()
a := NewTestAgent(t, t.Name(), TestACLConfig()+`
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
2018-10-19 16:04:07 +00:00
acl_master_token = ""
`)
defer a.Shutdown()
tests := []struct {
name string
method string
code int
token bool
}{
{"bootstrap", "PUT", http.StatusOK, true},
{"not again", "PUT", http.StatusForbidden, false},
}
testrpc.WaitForLeader(t, a.RPC, "dc1")
for _, tt := range tests {
t.Run(tt.name, func(t *testing.T) {
resp := httptest.NewRecorder()
req, _ := http.NewRequest(tt.method, "/v1/acl/bootstrap", nil)
out, err := a.srv.ACLBootstrap(resp, req)
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
2018-10-19 16:04:07 +00:00
if tt.token && err != nil {
t.Fatalf("err: %v", err)
}
if got, want := resp.Code, tt.code; got != want {
t.Fatalf("got %d want %d", got, want)
}
if tt.token {
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
2018-10-19 16:04:07 +00:00
wrap, ok := out.(*aclBootstrapResponse)
if !ok {
t.Fatalf("bad: %T", out)
}
if len(wrap.ID) != len("xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx") {
t.Fatalf("bad: %v", wrap)
}
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
2018-10-19 16:04:07 +00:00
if wrap.ID != wrap.SecretID {
t.Fatalf("bad: %v", wrap)
}
} else {
if out != nil {
t.Fatalf("bad: %T", out)
}
}
})
}
}
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
2018-10-19 16:04:07 +00:00
func TestACL_HTTP(t *testing.T) {
t.Parallel()
a := NewTestAgent(t, t.Name(), TestACLConfig())
2017-05-21 18:31:20 +00:00
defer a.Shutdown()
2014-08-06 17:30:47 +00:00
testrpc.WaitForLeader(t, a.RPC, "dc1")
2017-05-21 18:31:20 +00:00
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
2018-10-19 16:04:07 +00:00
idMap := make(map[string]string)
policyMap := make(map[string]*structs.ACLPolicy)
tokenMap := make(map[string]*structs.ACLToken)
// This is all done as a subtest for a couple reasons
// 1. It uses only 1 test agent and these are
// somewhat expensive to bring up and tear down often
// 2. Instead of having to bring up a new agent and prime
// the ACL system with some data before running the test
// we can intelligently order these tests so we can still
// test everything with less actual operations and do
// so in a manner that is less prone to being flaky
// 3. While this test will be large it should
t.Run("Policy", func(t *testing.T) {
t.Run("Create", func(t *testing.T) {
policyInput := &structs.ACLPolicy{
Name: "test",
Description: "test",
Rules: `acl = "read"`,
Datacenters: []string{"dc1"},
}
2017-05-21 18:31:20 +00:00
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
2018-10-19 16:04:07 +00:00
req, _ := http.NewRequest("PUT", "/v1/acl/policy?token=root", jsonBody(policyInput))
resp := httptest.NewRecorder()
obj, err := a.srv.ACLPolicyCreate(resp, req)
require.NoError(t, err)
policy, ok := obj.(*structs.ACLPolicy)
require.True(t, ok)
// 36 = length of the string form of uuids
require.Len(t, policy.ID, 36)
require.Equal(t, policyInput.Name, policy.Name)
require.Equal(t, policyInput.Description, policy.Description)
require.Equal(t, policyInput.Rules, policy.Rules)
require.Equal(t, policyInput.Datacenters, policy.Datacenters)
require.True(t, policy.CreateIndex > 0)
require.Equal(t, policy.CreateIndex, policy.ModifyIndex)
require.NotNil(t, policy.Hash)
require.NotEqual(t, policy.Hash, []byte{})
idMap["policy-test"] = policy.ID
policyMap[policy.ID] = policy
})
2014-08-06 17:30:47 +00:00
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
2018-10-19 16:04:07 +00:00
t.Run("Minimal", func(t *testing.T) {
policyInput := &structs.ACLPolicy{
Name: "minimal",
Rules: `key_prefix "" { policy = "read" }`,
}
2015-05-06 02:25:10 +00:00
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
2018-10-19 16:04:07 +00:00
req, _ := http.NewRequest("PUT", "/v1/acl/policy?token=root", jsonBody(policyInput))
resp := httptest.NewRecorder()
obj, err := a.srv.ACLPolicyCreate(resp, req)
require.NoError(t, err)
policy, ok := obj.(*structs.ACLPolicy)
require.True(t, ok)
// 36 = length of the string form of uuids
require.Len(t, policy.ID, 36)
require.Equal(t, policyInput.Name, policy.Name)
require.Equal(t, policyInput.Description, policy.Description)
require.Equal(t, policyInput.Rules, policy.Rules)
require.Equal(t, policyInput.Datacenters, policy.Datacenters)
require.True(t, policy.CreateIndex > 0)
require.Equal(t, policy.CreateIndex, policy.ModifyIndex)
require.NotNil(t, policy.Hash)
require.NotEqual(t, policy.Hash, []byte{})
idMap["policy-minimal"] = policy.ID
policyMap[policy.ID] = policy
})
2017-05-21 18:31:20 +00:00
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
2018-10-19 16:04:07 +00:00
t.Run("Name Chars", func(t *testing.T) {
policyInput := &structs.ACLPolicy{
Name: "read-all_nodes-012",
Rules: `node_prefix "" { policy = "read" }`,
}
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
2018-10-19 16:04:07 +00:00
req, _ := http.NewRequest("PUT", "/v1/acl/policy?token=root", jsonBody(policyInput))
resp := httptest.NewRecorder()
obj, err := a.srv.ACLPolicyCreate(resp, req)
require.NoError(t, err)
policy, ok := obj.(*structs.ACLPolicy)
require.True(t, ok)
// 36 = length of the string form of uuids
require.Len(t, policy.ID, 36)
require.Equal(t, policyInput.Name, policy.Name)
require.Equal(t, policyInput.Description, policy.Description)
require.Equal(t, policyInput.Rules, policy.Rules)
require.Equal(t, policyInput.Datacenters, policy.Datacenters)
require.True(t, policy.CreateIndex > 0)
require.Equal(t, policy.CreateIndex, policy.ModifyIndex)
require.NotNil(t, policy.Hash)
require.NotEqual(t, policy.Hash, []byte{})
idMap["policy-read-all-nodes"] = policy.ID
policyMap[policy.ID] = policy
})
2015-05-06 02:25:10 +00:00
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
2018-10-19 16:04:07 +00:00
t.Run("Update Name ID Mistmatch", func(t *testing.T) {
policyInput := &structs.ACLPolicy{
ID: "ac7560be-7f11-4d6d-bfcf-15633c2090fd",
Name: "read-all-nodes",
Description: "Can read all node information",
Rules: `node_prefix "" { policy = "read" }`,
Datacenters: []string{"dc1"},
}
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
2018-10-19 16:04:07 +00:00
req, _ := http.NewRequest("PUT", "/v1/acl/policy/"+idMap["policy-read-all-nodes"]+"?token=root", jsonBody(policyInput))
resp := httptest.NewRecorder()
_, err := a.srv.ACLPolicyCRUD(resp, req)
require.Error(t, err)
_, ok := err.(BadRequestError)
require.True(t, ok)
})
2017-05-21 18:31:20 +00:00
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
2018-10-19 16:04:07 +00:00
t.Run("Policy CRUD Missing ID in URL", func(t *testing.T) {
req, _ := http.NewRequest("GET", "/v1/acl/policy/?token=root", nil)
resp := httptest.NewRecorder()
_, err := a.srv.ACLPolicyCRUD(resp, req)
require.Error(t, err)
_, ok := err.(BadRequestError)
require.True(t, ok)
})
2014-08-06 17:30:47 +00:00
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
2018-10-19 16:04:07 +00:00
t.Run("Update", func(t *testing.T) {
policyInput := &structs.ACLPolicy{
Name: "read-all-nodes",
Description: "Can read all node information",
Rules: `node_prefix "" { policy = "read" }`,
Datacenters: []string{"dc1"},
}
2014-08-06 17:30:47 +00:00
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
2018-10-19 16:04:07 +00:00
req, _ := http.NewRequest("PUT", "/v1/acl/policy/"+idMap["policy-read-all-nodes"]+"?token=root", jsonBody(policyInput))
resp := httptest.NewRecorder()
obj, err := a.srv.ACLPolicyCRUD(resp, req)
require.NoError(t, err)
policy, ok := obj.(*structs.ACLPolicy)
require.True(t, ok)
// 36 = length of the string form of uuids
require.Len(t, policy.ID, 36)
require.Equal(t, policyInput.Name, policy.Name)
require.Equal(t, policyInput.Description, policy.Description)
require.Equal(t, policyInput.Rules, policy.Rules)
require.Equal(t, policyInput.Datacenters, policy.Datacenters)
require.True(t, policy.CreateIndex > 0)
require.True(t, policy.CreateIndex < policy.ModifyIndex)
require.NotNil(t, policy.Hash)
require.NotEqual(t, policy.Hash, []byte{})
idMap["policy-read-all-nodes"] = policy.ID
policyMap[policy.ID] = policy
})
2014-08-06 17:30:47 +00:00
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
2018-10-19 16:04:07 +00:00
t.Run("ID Supplied", func(t *testing.T) {
policyInput := &structs.ACLPolicy{
ID: "12123d01-37f1-47e6-b55b-32328652bd38",
Name: "with-id",
Description: "test",
Rules: `acl = "read"`,
Datacenters: []string{"dc1"},
}
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
2018-10-19 16:04:07 +00:00
req, _ := http.NewRequest("PUT", "/v1/acl/policy?token=root", jsonBody(policyInput))
resp := httptest.NewRecorder()
_, err := a.srv.ACLPolicyCreate(resp, req)
require.Error(t, err)
_, ok := err.(BadRequestError)
require.True(t, ok)
})
2017-05-21 18:31:20 +00:00
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
2018-10-19 16:04:07 +00:00
t.Run("Invalid payload", func(t *testing.T) {
body := bytes.NewBuffer(nil)
body.Write([]byte{0, 1, 2, 3, 4, 5, 6, 7, 8, 9})
2014-08-06 17:30:47 +00:00
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
2018-10-19 16:04:07 +00:00
req, _ := http.NewRequest("PUT", "/v1/acl/policy?token=root", body)
resp := httptest.NewRecorder()
_, err := a.srv.ACLPolicyCreate(resp, req)
require.Error(t, err)
_, ok := err.(BadRequestError)
require.True(t, ok)
})
t.Run("Delete", func(t *testing.T) {
req, _ := http.NewRequest("DELETE", "/v1/acl/policy/"+idMap["policy-minimal"]+"?token=root", nil)
resp := httptest.NewRecorder()
_, err := a.srv.ACLPolicyCRUD(resp, req)
require.NoError(t, err)
delete(policyMap, idMap["policy-minimal"])
delete(idMap, "policy-minimal")
})
t.Run("List", func(t *testing.T) {
req, _ := http.NewRequest("GET", "/v1/acl/policies?token=root", nil)
resp := httptest.NewRecorder()
raw, err := a.srv.ACLPolicyList(resp, req)
require.NoError(t, err)
policies, ok := raw.(structs.ACLPolicyListStubs)
require.True(t, ok)
// 2 we just created + global management
require.Len(t, policies, 3)
for policyID, expected := range policyMap {
found := false
for _, actual := range policies {
if actual.ID == policyID {
require.Equal(t, expected.Name, actual.Name)
require.Equal(t, expected.Datacenters, actual.Datacenters)
require.Equal(t, expected.Hash, actual.Hash)
require.Equal(t, expected.CreateIndex, actual.CreateIndex)
require.Equal(t, expected.ModifyIndex, actual.ModifyIndex)
found = true
break
}
}
require.True(t, found)
}
})
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
2018-10-19 16:04:07 +00:00
t.Run("Read", func(t *testing.T) {
req, _ := http.NewRequest("GET", "/v1/acl/policy/"+idMap["policy-read-all-nodes"]+"?token=root", nil)
resp := httptest.NewRecorder()
raw, err := a.srv.ACLPolicyCRUD(resp, req)
require.NoError(t, err)
policy, ok := raw.(*structs.ACLPolicy)
require.True(t, ok)
require.Equal(t, policyMap[idMap["policy-read-all-nodes"]], policy)
})
2014-08-06 17:30:47 +00:00
})
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
2018-10-19 16:04:07 +00:00
t.Run("Token", func(t *testing.T) {
t.Run("Create", func(t *testing.T) {
tokenInput := &structs.ACLToken{
Description: "test",
Policies: []structs.ACLTokenPolicyLink{
structs.ACLTokenPolicyLink{
ID: idMap["policy-test"],
Name: policyMap[idMap["policy-test"]].Name,
},
structs.ACLTokenPolicyLink{
ID: idMap["policy-read-all-nodes"],
Name: policyMap[idMap["policy-read-all-nodes"]].Name,
},
},
}
2014-08-06 17:30:47 +00:00
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
2018-10-19 16:04:07 +00:00
req, _ := http.NewRequest("PUT", "/v1/acl/token?token=root", jsonBody(tokenInput))
resp := httptest.NewRecorder()
obj, err := a.srv.ACLTokenCreate(resp, req)
require.NoError(t, err)
token, ok := obj.(*structs.ACLToken)
require.True(t, ok)
// 36 = length of the string form of uuids
require.Len(t, token.AccessorID, 36)
require.Len(t, token.SecretID, 36)
require.Equal(t, tokenInput.Description, token.Description)
require.Equal(t, tokenInput.Policies, token.Policies)
require.True(t, token.CreateIndex > 0)
require.Equal(t, token.CreateIndex, token.ModifyIndex)
require.NotNil(t, token.Hash)
require.NotEqual(t, token.Hash, []byte{})
idMap["token-test"] = token.AccessorID
tokenMap[token.AccessorID] = token
})
t.Run("Create Local", func(t *testing.T) {
tokenInput := &structs.ACLToken{
Description: "local",
Policies: []structs.ACLTokenPolicyLink{
structs.ACLTokenPolicyLink{
ID: idMap["policy-test"],
Name: policyMap[idMap["policy-test"]].Name,
},
structs.ACLTokenPolicyLink{
ID: idMap["policy-read-all-nodes"],
Name: policyMap[idMap["policy-read-all-nodes"]].Name,
},
},
Local: true,
}
2014-08-12 21:48:36 +00:00
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
2018-10-19 16:04:07 +00:00
req, _ := http.NewRequest("PUT", "/v1/acl/token?token=root", jsonBody(tokenInput))
resp := httptest.NewRecorder()
obj, err := a.srv.ACLTokenCreate(resp, req)
require.NoError(t, err)
token, ok := obj.(*structs.ACLToken)
require.True(t, ok)
// 36 = length of the string form of uuids
require.Len(t, token.AccessorID, 36)
require.Len(t, token.SecretID, 36)
require.Equal(t, tokenInput.Description, token.Description)
require.Equal(t, tokenInput.Policies, token.Policies)
require.True(t, token.CreateIndex > 0)
require.Equal(t, token.CreateIndex, token.ModifyIndex)
require.NotNil(t, token.Hash)
require.NotEqual(t, token.Hash, []byte{})
idMap["token-local"] = token.AccessorID
tokenMap[token.AccessorID] = token
})
t.Run("Read", func(t *testing.T) {
expected := tokenMap[idMap["token-test"]]
req, _ := http.NewRequest("GET", "/v1/acl/token/"+expected.AccessorID+"?token=root", nil)
resp := httptest.NewRecorder()
obj, err := a.srv.ACLTokenCRUD(resp, req)
require.NoError(t, err)
token, ok := obj.(*structs.ACLToken)
require.True(t, ok)
require.Equal(t, expected, token)
})
t.Run("Self", func(t *testing.T) {
expected := tokenMap[idMap["token-test"]]
req, _ := http.NewRequest("GET", "/v1/acl/token/self?token="+expected.SecretID, nil)
resp := httptest.NewRecorder()
obj, err := a.srv.ACLTokenSelf(resp, req)
require.NoError(t, err)
token, ok := obj.(*structs.ACLToken)
require.True(t, ok)
require.Equal(t, expected, token)
})
t.Run("Clone", func(t *testing.T) {
tokenInput := &structs.ACLToken{
Description: "cloned token",
}
2017-05-21 18:31:20 +00:00
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
2018-10-19 16:04:07 +00:00
baseToken := tokenMap[idMap["token-test"]]
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
2018-10-19 16:04:07 +00:00
req, _ := http.NewRequest("PUT", "/v1/acl/token/"+baseToken.AccessorID+"/clone?token=root", jsonBody(tokenInput))
resp := httptest.NewRecorder()
obj, err := a.srv.ACLTokenCRUD(resp, req)
require.NoError(t, err)
token, ok := obj.(*structs.ACLToken)
require.True(t, ok)
require.NotEqual(t, baseToken.AccessorID, token.AccessorID)
require.NotEqual(t, baseToken.SecretID, token.SecretID)
require.Equal(t, tokenInput.Description, token.Description)
require.Equal(t, baseToken.Policies, token.Policies)
require.True(t, token.CreateIndex > 0)
require.Equal(t, token.CreateIndex, token.ModifyIndex)
require.NotNil(t, token.Hash)
require.NotEqual(t, token.Hash, []byte{})
idMap["token-cloned"] = token.AccessorID
tokenMap[token.AccessorID] = token
})
t.Run("Update", func(t *testing.T) {
originalToken := tokenMap[idMap["token-cloned"]]
// Accessor and Secret will be filled in
tokenInput := &structs.ACLToken{
Description: "Better description for this cloned token",
Policies: []structs.ACLTokenPolicyLink{
structs.ACLTokenPolicyLink{
ID: idMap["policy-read-all-nodes"],
Name: policyMap[idMap["policy-read-all-nodes"]].Name,
},
},
}
2017-05-21 18:31:20 +00:00
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
2018-10-19 16:04:07 +00:00
req, _ := http.NewRequest("PUT", "/v1/acl/token/"+originalToken.AccessorID+"?token=root", jsonBody(tokenInput))
resp := httptest.NewRecorder()
obj, err := a.srv.ACLTokenCRUD(resp, req)
require.NoError(t, err)
token, ok := obj.(*structs.ACLToken)
require.True(t, ok)
require.Equal(t, originalToken.AccessorID, token.AccessorID)
require.Equal(t, originalToken.SecretID, token.SecretID)
require.Equal(t, tokenInput.Description, token.Description)
require.Equal(t, tokenInput.Policies, token.Policies)
require.True(t, token.CreateIndex > 0)
require.True(t, token.CreateIndex < token.ModifyIndex)
require.NotNil(t, token.Hash)
require.NotEqual(t, token.Hash, []byte{})
require.NotEqual(t, token.Hash, originalToken.Hash)
tokenMap[token.AccessorID] = token
})
t.Run("CRUD Missing Token Accessor ID", func(t *testing.T) {
req, _ := http.NewRequest("GET", "/v1/acl/token/?token=root", nil)
resp := httptest.NewRecorder()
obj, err := a.srv.ACLTokenCRUD(resp, req)
require.Error(t, err)
require.Nil(t, obj)
_, ok := err.(BadRequestError)
require.True(t, ok)
})
t.Run("Update Accessor Mismatch", func(t *testing.T) {
originalToken := tokenMap[idMap["token-cloned"]]
// Accessor and Secret will be filled in
tokenInput := &structs.ACLToken{
AccessorID: "e8aeb69a-0ace-42b9-b95f-d1d9eafe1561",
Description: "Better description for this cloned token",
Policies: []structs.ACLTokenPolicyLink{
structs.ACLTokenPolicyLink{
ID: idMap["policy-read-all-nodes"],
Name: policyMap[idMap["policy-read-all-nodes"]].Name,
},
},
}
req, _ := http.NewRequest("PUT", "/v1/acl/token/"+originalToken.AccessorID+"?token=root", jsonBody(tokenInput))
resp := httptest.NewRecorder()
obj, err := a.srv.ACLTokenCRUD(resp, req)
require.Error(t, err)
require.Nil(t, obj)
_, ok := err.(BadRequestError)
require.True(t, ok)
})
t.Run("Delete", func(t *testing.T) {
req, _ := http.NewRequest("DELETE", "/v1/acl/token/"+idMap["token-cloned"]+"?token=root", nil)
resp := httptest.NewRecorder()
_, err := a.srv.ACLTokenCRUD(resp, req)
require.NoError(t, err)
delete(tokenMap, idMap["token-cloned"])
delete(idMap, "token-cloned")
})
t.Run("List", func(t *testing.T) {
req, _ := http.NewRequest("GET", "/v1/acl/tokens?token=root", nil)
resp := httptest.NewRecorder()
raw, err := a.srv.ACLTokenList(resp, req)
require.NoError(t, err)
tokens, ok := raw.(structs.ACLTokenListStubs)
require.True(t, ok)
// 3 tokens created but 1 was deleted + master token + anon token
require.Len(t, tokens, 4)
// this loop doesn't verify anything about the master token
for tokenID, expected := range tokenMap {
found := false
for _, actual := range tokens {
if actual.AccessorID == tokenID {
require.Equal(t, expected.Description, actual.Description)
require.Equal(t, expected.Policies, actual.Policies)
require.Equal(t, expected.Local, actual.Local)
require.Equal(t, expected.CreateTime, actual.CreateTime)
require.Equal(t, expected.Hash, actual.Hash)
require.Equal(t, expected.CreateIndex, actual.CreateIndex)
require.Equal(t, expected.ModifyIndex, actual.ModifyIndex)
found = true
break
}
}
require.True(t, found)
}
})
t.Run("List by Policy", func(t *testing.T) {
req, _ := http.NewRequest("GET", "/v1/acl/tokens?token=root&policy="+structs.ACLPolicyGlobalManagementID, nil)
resp := httptest.NewRecorder()
raw, err := a.srv.ACLTokenList(resp, req)
require.NoError(t, err)
tokens, ok := raw.(structs.ACLTokenListStubs)
require.True(t, ok)
require.Len(t, tokens, 1)
token := tokens[0]
require.Equal(t, "Master Token", token.Description)
require.Len(t, token.Policies, 1)
require.Equal(t, structs.ACLPolicyGlobalManagementID, token.Policies[0].ID)
})
})
}