* A few API mods and unit tests.
* Update the unit tests to verify query/write metadata and to fix the rules endpoint tests.
* Make sure the full information for the replication status is in the api packge
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.
* agent/debug: add package for debugging, host info
* api: add v1/agent/host endpoint
* agent: add v1/agent/host endpoint
* command/debug: implementation of static capture
* command/debug: tests and only configured targets
* agent/debug: add basic test for host metrics
* command/debug: add methods for dynamic data capture
* api: add debug/pprof endpoints
* command/debug: add pprof
* command/debug: timing, wg, logs to disk
* vendor: add gopsutil/disk
* command/debug: add a usage section
* website: add docs for consul debug
* agent/host: require operator:read
* api/host: improve docs and no retry timing
* command/debug: fail on extra arguments
* command/debug: fixup file permissions to 0644
* command/debug: remove server flags
* command/debug: improve clarity of usage section
* api/debug: add Trace for profiling, fix profile
* command/debug: capture profile and trace at the same time
* command/debug: add index document
* command/debug: use "clusters" in place of members
* command/debug: remove address in output
* command/debug: improve comment on metrics sleep
* command/debug: clarify usage
* agent: always register pprof handlers and protect
This will allow us to avoid a restart of a target agent
for profiling by always registering the pprof handlers.
Given this is a potentially sensitive path, it is protected
with an operator:read ACL and enable debug being
set to true on the target agent. enable_debug still requires
a restart.
If ACLs are disabled, enable_debug is sufficient.
* command/debug: use trace.out instead of .prof
More in line with golang docs.
* agent: fix comment wording
* agent: wrap table driven tests in t.run()
The default http.Client uses infinite timeouts, so if TestHTTPAPI_MethodNotAllowed_OSS experienced anything going wrong about setup it could hang forever.
Switching to hard coding various http.Client timeouts to non-infinite values at least bounds the failure time.
* Add -enable-local-script-checks options
These options allow for a finer control over when script checks are enabled by
giving the option to only allow them when they are declared from the local
file system.
* Add documentation for the new option
* Nitpick doc wording
* Support multiple tags for health and catalog api endpoints
Fixes#1781.
Adds a `ServiceTags` field to the ServiceSpecificRequest to support
multiple tags, updates the filter logic in the catalog store, and
propagates these change through to the health and catalog endpoints.
Note: Leaves `ServiceTag` in the struct, since it is being used as
part of the DNS lookup, which in turn uses the health check.
* Update the api package to support multiple tags
Includes additional tests.
* Update new tests to use the `require` library
* Update HealthConnect check after a bad merge
* [Performance On Large clusters] Checks do update services/nodes only when really modified to avoid too many updates on very large clusters
In a large cluster, when having a few thousands of nodes, the anti-entropy
mechanism performs lots of changes (several per seconds) while
there is no real change. This patch wants to improve this in order
to increase Consul scalability when using many blocking requests on
health for instance.
* [Performance for large clusters] Only updates index of service if service is really modified
* [Performance for large clusters] Only updates index of nodes if node is really modified
* Added comments / ensure IsSame() has clear semantics
* Avoid having modified boolean, return nil directly if stutures are Same
* Fixed unstable unit tests TestLeader_ChangeServerID
* Rewrite TestNode_IsSame() for better readability as suggested by @banks
* Rename ServiceNode.IsSame() into IsSameService() + added unit tests
* Do not duplicate TestStructs_ServiceNode_Conversions() and increase test coverage of IsSameService
* Clearer documentation in IsSameService
* Take into account ServiceProxy into ServiceNode.IsSameService()
* Fixed IsSameService() with all new structures
* Initial draft of Sidecar Service and Managed Proxy deprecation docs
* Service definition deprecation notices and sidecar service
* gRPC and sidecar service config options; Deprecate managed proxy options
* Envoy Docs: Basic envoy command; envoy getting started/intro
* Remove change that snuck in
* Envoy custom config example
* Add agent/service API docs; deprecate proxy config endpoint
* Misc grep cleanup for managed proxies; capitalize Envoy
* Updates to getting started guide
* Add missing link
* Refactor Envoy guide into a separate guide and add bootstrap reference notes.
* Add limitations to Envoy docs; Highlight no fixes for known managed proxy issues on deprecation page; clarify snake cae stuff; Sidecar Service lifecycle
* Fix bug in leaf-cert cache type where multiple clients with different tokens would share certs and block incorrectly
* Use hash for issued certs key to avoid ambiguity concatenating
* Plumb xDS server and proxyxfg into the agent startup
* Add `consul connect envoy` command to allow running Envoy as a connect sidecar.
* Add test for help tabs; typos and style fixups from review
* Vendor updates for gRPC and xDS server
* xDS server implementation for serving Envoy as a Connect proxy
* Address initial review comments
* consistent envoy package aliases; typos fixed; override TLS and authz for custom listeners
* Moar Typos
* Moar typos
* Proxy Config Manager
This component watches for local state changes on the agent and ensures that each service registered locally with Kind == connect-proxy has it's state being actively populated in the cache.
This serves two purposes:
1. For the built-in proxy, it ensures that the state needed to accept connections is available in RAM shortly after registration and likely before the proxy actually starts accepting traffic.
2. For (future - next PR) xDS server and other possible future proxies that require _push_ based config discovery, this provides a mechanism to subscribe and be notified about updates to a proxy instance's config including upstream service discovery results.
* Address review comments
* Better comments; Better delivery of latest snapshot for slow watchers; Embed Config
* Comment typos
* Add upstream Stringer for funsies
- A new endpoint `/v1/agent/service/:service_id` which is a generic way to look up the service for a single instance. The primary value here is that it:
- **supports hash-based blocking** and so;
- **replaces `/agent/connect/proxy/:proxy_id`** as the mechanism the built-in proxy uses to read its config.
- It's not proxy specific and so works for any service.
- It has a temporary shim to call through to the existing endpoint to preserve current managed proxy config defaulting behaviour until that is removed entirely (tested).
- The built-in proxy now uses the new endpoint exclusively for it's config
- The built-in proxy now has a `-sidecar-for` flag that allows the service ID of the _target_ service to be specified, on the condition that there is exactly one "sidecar" proxy (that is one that has `Proxy.DestinationServiceID` set) for the service registered.
- Several fixes for edge cases for SidecarService
- A fix for `Alias` checks - when running locally they didn't update their state until some external thing updated the target. If the target service has no checks registered as below, then the alias never made it past critical.
* Added new Config for SidecarService in ServiceDefinitions.
* WIP: all the code needed for SidecarService is written... none of it is tested other than config :). Need API updates too.
* Test coverage for the new sidecarServiceFromNodeService method.
* Test API registratrion with SidecarService
* Recursive Key Translation 🤦
* Add tests for nested sidecar defintion arrays to ensure they are translated correctly
* Use dedicated internal state rather than Service Meta for tracking sidecars for deregistration.
Add tests for deregistration.
* API struct for agent register. No other endpoint should be affected yet.
* Additional test cases to cover updates to API registrations
* Refactor Service Definition ProxyDestination.
This includes:
- Refactoring all internal structs used
- Updated tests for both deprecated and new input for:
- Agent Services endpoint response
- Agent Service endpoint response
- Agent Register endpoint
- Unmanaged deprecated field
- Unmanaged new fields
- Managed deprecated upstreams
- Managed new
- Catalog Register
- Unmanaged deprecated field
- Unmanaged new fields
- Managed deprecated upstreams
- Managed new
- Catalog Services endpoint response
- Catalog Node endpoint response
- Catalog Service endpoint response
- Updated API tests for all of the above too (both deprecated and new forms of register)
TODO:
- config package changes for on-disk service definitions
- proxy config endpoint
- built-in proxy support for new fields
* Agent proxy config endpoint updated with upstreams
* Config file changes for upstreams.
* Add upstream opaque config and update all tests to ensure it works everywhere.
* Built in proxy working with new Upstreams config
* Command fixes and deprecations
* Fix key translation, upstream type defaults and a spate of other subtele bugs found with ned to end test scripts...
TODO: tests still failing on one case that needs a fix. I think it's key translation for upstreams nested in Managed proxy struct.
* Fix translated keys in API registration.
≈
* Fixes from docs
- omit some empty undocumented fields in API
- Bring back ServiceProxyDestination in Catalog responses to not break backwards compat - this was removed assuming it was only used internally.
* Documentation updates for Upstreams in service definition
* Fixes for tests broken by many refactors.
* Enable travis on f-connect branch in this branch too.
* Add consistent Deprecation comments to ProxyDestination uses
* Update version number on deprecation notices, and correct upstream datacenter field with explanation in docs
* Rename agent/proxy package to reflect that it is limited to managed proxy processes
Rationale: we have several other components of the agent that relate to Connect proxies for example the ProxyConfigManager component needed for Envoy work. Those things are pretty separate from the focus of this package so far which is only concerned with managing external proxy processes so it's nota good fit to put code for that in here, yet there is a naming clash if we have other packages related to proxy functionality that are not in the `agent/proxy` package.
Happy to bikeshed the name. I started by calling it `managedproxy` but `managedproxy.Manager` is especially unpleasant. `proxyprocess` seems good in that it's more specific about purpose but less clearly connected with the concept of "managed proxies". The names in use are cleaner though e.g. `proxyprocess.Manager`.
This rename was completed automatically using golang.org/x/tools/cmd/gomvpkg.
Depends on #4541
* Fix missed windows tagged files
* Add cache types for catalog/services and health/services and basic test that caching works
* Support non-blocking cache types with Cache-Control semantics.
* Update API docs to include caching info for every endpoint.
* Comment updates per PR feedback.
* Add note on caching to the 10,000 foot view on the architecture page to make the new data path more clear.
* Document prepared query staleness quirk and force all background requests to AllowStale so we can spread service discovery load across servers.
Don't set the value of TLSConfig.Address explicitly.
This will make sure env vars like CONSUL_TLS_SERVER_NAME are taken into account for the connection. Fixes#4718.
In a real agent the `cache` instance is alive until the agent shuts down so this is not a real leak in production, however in out test suite, every testAgent that is started and stops leaks goroutines that never get cleaned up which accumulate consuming CPU and memory through subsequent test in the `agent` package which doesn't help our test flakiness.
This adds a Close method that doesn't invalidate or clean up the cache, and still allows concurrent blocking queries to run (for up to 10 mins which might still affect tests). But at least it doesn't maintain them forever with background refresh and an expiry watcher routine.
It would be nice to cancel any outstanding blocking requests as well when we close but that requires much more invasive surgery right into our RPC protocol since we don't have a way to cancel requests currently.
Unscientifically this seems to make tests pass a bit quicker and more reliably locally but I can't really be sure of that!
cli: forward SIGTERM to child process of 'lock' and 'watch' subcommands on unix
This also removes the signal handler for SIGKILL as it's impossible to receive these signals.
* Fix CA pruning when CA config uses string durations.
The tl;dr here is:
- Configuring LeafCertTTL with a string like "72h" is how we do it by default and should be supported
- Most of our tests managed to escape this by defining them as time.Duration directly
- Out actual default value is a string
- Since this is stored in a map[string]interface{} config, when it is written to Raft it goes through a msgpack encode/decode cycle (even though it's written from server not over RPC).
- msgpack decode leaves the string as a `[]uint8`
- Some of our parsers required string and failed
- So after 1 hour, a default configured server would throw an error about pruning old CAs
- If a new CA was configured that set LeafCertTTL as a time.Duration, things might be OK after that, but if a new CA was just configured from config file, intialization would cause same issue but always fail still so would never prune the old CA.
- Mostly this is just a janky error that got passed tests due to many levels of complicated encoding/decoding.
tl;dr of the tl;dr: Yay for type safety. Map[string]interface{} combined with msgpack always goes wrong but we somehow get bitten every time in a new way :D
We already fixed this once! The main CA config had the same problem so @kyhavlov already wrote the mapstructure DecodeHook that fixes it. It wasn't used in several places it needed to be and one of those is notw in `structs` which caused a dependency cycle so I've moved them.
This adds a whole new test thta explicitly tests the case that broke here. It also adds tests that would have failed in other places before (Consul and Vaul provider parsing functions). I'm not sure if they would ever be affected as it is now as we've not seen things broken with them but it seems better to explicitly test that and support it to not be bitten a third time!
* Typo fix
* Fix bad Uint8 usage
If you provide an invalid HTTP configuration consul will still start again instead of failing. But if you do so the build-in proxy won't be able to start which you might need for connect.
This implements parts of RFC 7871 where Consul is acting as an authoritative name server (or forwarding resolver when recursors are configured)
If ECS opt is present in the request we will mirror it back and return a response with a scope of 0 (global) or with the same prefix length as the request (indicating its valid specifically for that subnet).
We only mirror the prefix-length (non-global) for prepared queries as those could potentially use nearness checks that could be affected by the subnet. In the future we could get more sophisticated with determining the scope bits and allow for better caching of prepared queries that don’t rely on nearness checks.
The other thing this does not do is implement the part of the ECS RFC related to originating ECS headers when acting as a intermediate DNS server (forwarding resolver). That would take a quite a bit more effort and in general provide very little value. Consul will currently forward the ECS headers between recursors and the clients transparently, we just don't originate them for non-ECS clients to get potentially more accurate "location aware" results.
Fixes: #4578
Prior to this fix if there was an error binding to ports for the DNS servers the error would be swallowed by the gated log writer and never output. This fix propagates the DNS server errors back to the shell with a multierror.
* Implementation of Weights Data structures
Adding this datastructure will allow us to resolve the
issues #1088 and #4198
This new structure defaults to values:
```
{ Passing: 1, Warning: 0 }
```
Which means, use weight of 0 for a Service in Warning State
while use Weight 1 for a Healthy Service.
Thus it remains compatible with previous Consul versions.
* Implemented weights for DNS SRV Records
* DNS properly support agents with weight support while server does not (backwards compatibility)
* Use Warning value of Weights of 1 by default
When using DNS interface with only_passing = false, all nodes
with non-Critical healthcheck used to have a weight value of 1.
While having weight.Warning = 0 as default value, this is probably
a bad idea as it breaks ascending compatibility.
Thus, we put a default value of 1 to be consistent with existing behaviour.
* Added documentation for new weight field in service description
* Better documentation about weights as suggested by @banks
* Return weight = 1 for unknown Check states as suggested by @banks
* Fixed typo (of -> or) in error message as requested by @mkeeler
* Fixed unstable unit test TestRetryJoin
* Fixed unstable tests
* Fixed wrong Fatalf format in `testrpc/wait.go`
* Added notes regarding DNS SRV lookup limitations regarding number of instances
* Documentation fixes and clarification regarding SRV records with weights as requested by @banks
* Rephrase docs
* Added log-file flag to capture Consul logs in a user specified file
* Refactored code.
* Refactored code. Added flags to rotate logs based on bytes and duration
* Added the flags for log file and log rotation on the webpage
* Fixed TestSantize from failing due to the addition of 3 flags
* Introduced changes : mutex, data-dir log writes, rotation logic
* Added test for logfile and updated the default log destination for docs
* Log name now uses UnixNano
* TestLogFile is now uses t.Parallel()
* Removed unnecessary int64Val function
* Updated docs to reflect default log name for log-file
* No longer writes to data-dir and adds .log if the filename has no extension
Fixes#4515
This just slightly refactors the logic to only attempt to set the serf wan reconnect timeout when the rest of the serf wan settings are configured - thus avoiding a segfault.
* Display more information about check being not properly added when it fails
It follows an incident where we add lots of error messages:
[WARN] consul.fsm: EnsureRegistration failed: failed inserting check: Missing service registration
That seems related to Consul failing to restart on respective agents.
Having Node information as well as service information would help diagnose the issue.
* Renamed ensureCheckIfNodeMatches() as requested by @banks
- Add WaitForTestAgent to tests flaky due to missing serfHealth registration
- Fix bug in retries calling Fatalf with *testing.T
- Convert TestLockCommand_ChildExitCode to table driven test
* Allow to rename nodes with IDs, will fix#3974 and #4413
This change allow to rename any well behaving recent agent with an
ID to be renamed safely, ie: without taking the name of another one
with case insensitive comparison.
Deprecated behaviour warning
----------------------------
Due to asceding compatibility, it is still possible however to
"take" the name of another name by not providing any ID.
Note that when not providing any ID, it is possible to have 2 nodes
having similar names with case differences, ie: myNode and mynode
which might lead to DB corruption on Consul server side and
lead to server not properly restarting.
See #3983 and #4399 for Context about this change.
Disabling registration of nodes without IDs as specified in #4414
should probably be the way to go eventually.
* Removed the case-insensitive search when adding a node within the else
block since it breaks the test TestAgentAntiEntropy_Services
While the else case is probably legit, it will be fixed with #4414 in
a later release.
* Added again the test in the else to avoid duplicated names, but
enforce this test only for nodes having IDs.
Thus most tests without any ID will work, and allows us fixing
* Added more tests regarding request with/without IDs.
`TestStateStore_EnsureNode` now test registration and renaming with IDs
`TestStateStore_EnsureNodeDeprecated` tests registration without IDs
and tests removing an ID from a node as well as updated a node
without its ID (deprecated behaviour kept for backwards compatibility)
* Do not allow renaming in case of conflict, including when other node has no ID
* Fixed function GetNodeID that was not working due to wrong type when searching node from its ID
Thus, all tests about renaming were not working properly.
Added the full test cas that allowed me to detect it.
* Better error messages, more tests when nodeID is not a valid UUID in GetNodeID()
* Added separate TestStateStore_GetNodeID to test GetNodeID.
More complete test coverage for GetNodeID
* Added new unit test `TestStateStore_ensureNoNodeWithSimilarNameTxn`
Also fixed comments to be clearer after remarks from @banks
* Fixed error message in unit test to match test case
* Use uuid.ParseUUID to parse Node.ID as requested by @mkeeler
* Fixes TestAgent_IndexChurn
* Fixes TestPreparedQuery_Wrapper
* Increased sleep in agent_test for IndexChurn to 500ms
* Made the comment about joinWAN operation much less of a cliffhanger
* Revert "BUGFIX: Unit test relying on WaitForLeader() did not work due to wrong test (#4472)"
This reverts commit cec5d7239621e0732b3f70158addb1899442acb3.
* Revert "CA initialization while boostrapping and TestLeader_ChangeServerID fix. (#4493)"
This reverts commit 589b589b53e56af38de25db9b56967bdf1f2c069.
- Improve resilience of testrpc.WaitForLeader()
- Add additionall retry to CI
- Increase "go test" timeout to 8m
- Add wait for cluster leader to several tests in the agent package
- Add retry to some tests in the api and command packages
* Fixes the DNS recursor properly resolving the requests
* Added a test case for the recursor bug
* Refactored code && added a test case for all failing recursors
* Inner indentation moved into else if check
Fixes: #4441
This fixes the issue with Connect Managed Proxies + ACLs being broken.
The underlying problem was that the token parsed for most http endpoints was sent untouched to the servers via the RPC request. These changes make it so that at the HTTP endpoint when parsing the token we additionally attempt to convert potential proxy tokens into regular tokens before sending to the RPC endpoint. Proxy tokens are only valid on the agent with the managed proxy so the resolution has to happen before it gets forwarded anywhere.
* New Providers added and updated vendoring for go-discover
* Vendor.json formatted using make vendorfmt
* Docs/Agent/auto-join: Added documentation for the new providers introduced in this PR
* Updated the golang.org/x/sys/unix in the vendor directory
* Agent: TestGoDiscoverRegistration updated to reflect the addition of new providers
* Deleted terraform.tfstate from vendor.
* Deleted terraform.tfstate.backup
Deleted terraform state file artifacts from unknown runs.
* Updated x/sys/windows vendor for Windows binary compilation
* Fix theoretical cache collision bug if/when we use more cache types with same result type
* Generalized fix for blocking query handling when state store methods return zero index
* Refactor test retry to only affect CI
* Undo make file merge
* Add hint to error message returned to end-user requests if Connect is not enabled when they try to request cert
* Explicit error for Roots endpoint if connect is disabled
* Fix tests that were asserting old behaviour
Also change how loadProxies works. Now it will load all persisted proxies into a map, then when loading config file proxies will look up the previous proxy token in that map.
UUID auto-generation here causes trouble in a few cases. The biggest being older
nodes reregistering will fail when the UUIDs are different and the names match
This reverts commit 0f700340828f464449c2e0d5a82db0bc5456d385.
This reverts commit d1a8f9cb3f6f48dd9c8d0bc858031ff6ccff51d0.
This reverts commit cf69ec42a418ab6594a6654e9545e12160f30970.
Since DNS is case insensitive and DB as issues when similar names with different
cases are added, check for unicity based on case insensitivity.
Following another big incident we had in our cluster, we also validate
that adding/renaming a not does not conflicts with case insensitive
matches.
We had the following error once:
- one node called: mymachine.MYDC.mydomain was shut off
- another node (different ID) was added with name: mymachine.mydc.mydomain before
72 hours
When restarting the consul server of domain, the consul server restarted failed
to start since it detected an issue in RAFT database because
mymachine.MYDC.mydomain and mymachine.mydc.mydomain had the same names.
Checking at registration time with case insensitivity should definitly fix
those issues and avoid Consul DB corruption.
It fixes the following warnings:
agent/config/builder.go:1201: Errorf format %q has arg s of wrong type *string
agent/config/builder.go:1240: Errorf format %q has arg s of wrong type *string
This also changes where the enforcement of the enable_additional_node_meta_txt configuration gets applied.
formatNodeRecord returns the main RRs and the meta/TXT RRs in separate slices. Its then up to the caller to add to the appropriate sections or not.
This just makes sure that if multiple services are registered with unique service addresses that we don’t blast back multiple CNAMEs for the same service DNS name and keeps us within the DNS specs.
It will allow the following:
* when connectivity is limited (saturated linnks between DCs), only one
single request to refresh an ACL will be sent to ACL master DC instead
of statcking ACL refresh queries
* when extend-cache is used for ACL, do not wait for result, but refresh
the ACL asynchronously, so no delay is not impacting slave DC
* When extend-cache is not used, keep the existing blocking mechanism,
but only send a single refresh request.
This will fix https://github.com/hashicorp/consul/issues/3524
These were only added as SPIFFE intends to use the in the future but currently does not mandate their usage due to patch support in common TLS implementations and some ambiguity over how to use them with URI SAN certificates. We included them because until now everything seem fine with it, however we've found the latest version of `openssl` (1.1.0h) fails to validate our certificats if its enabled. LibreSSL as installed on OS X by default doesn’t have these issues. For now it's most compatible not to have them and later we can find ways to add constraints with wider compatibility testing.
Few other fixes in here just to get a clean run locally - they are all also fixed in other PRs but shouldn't conflict.
This should be robust to timing between goroutines now.
- Dev mode assumed no persistence of services although proxy state is persisted which caused proxies to be killed on startup as their services were no longer registered. Fixed.
- Didn't snapshot the ProxyID which meant that proxies were adopted OK from snapshot but failed to restart if they died since there was no proxyID in the ENV on restart
- Dev mode with no persistence just kills all proxies on shutdown since it can't recover them later
- Naming things
This turns out to have a lot more subtelty than we accounted for. The test suite is especially prone to races now we can only poll the child and many extra levels of indirectoin are needed to correctly run daemon process without it becoming a Zombie.
I ran this test suite in a loop with parallel enabled to verify for races (-race doesn't find any as they are logical inter-process ones not actual data races). I made it through ~50 runs before hitting an error due to timing which is much better than before. I want to go back and see if we can do better though. Just getting this up.
Kyle Havlovitz