Fix a logmon leak causing high goroutine and memory usage when a task
restarts.
Logmon `FileRotator` buffers the task stdout/stderr streams and
periodically flushing them to log files. Logmon creates a new
FileRotator for each stream for each task run. However, the
`flushPeriodically` goroutine is leaked when a task restarts,
holding a reference to a no-longer-needed `FileRotator` instance
along with its 64kb buffer.
The cause is that the code assumed `time.Ticker.Stop()` closes the
ticker channel, thereby terminating the goroutine, but the documentation
says otherwise:
> Stop turns off a ticker. After Stop, no more ticks will be sent. Stop does not close the channel, to prevent a concurrent goroutine reading from the channel from seeing an erroneous "tick".
https://pkg.go.dev/time#Ticker.Stop
By default we should not expose the NOMAD_LICENSE environment variable
to tasks.
Also refactor where the DefaultEnvDenyList lives so we don't have to
maintain 2 copies of it. Since client/config is the most obvious
location, keep a reference there to its unfortunate home buried deep
in command/agent/host. Since the agent uses this list as well for the
/agent/host endpoint the list must be accessible from both command/agent
and client.
Add a new hostname string parameter to the network block which
allows operators to specify the hostname of the network namespace.
Changing this causes a destructive update to the allocation and it
is omitted if empty from API responses. This parameter also supports
interpolation.
In order to have a hostname passed as a configuration param when
creating an allocation network, the CreateNetwork func of the
DriverNetworkManager interface needs to be updated. In order to
minimize the disruption of future changes, rather than add another
string func arg, the function now accepts a request struct along with
the allocID param. The struct has the hostname as a field.
The in-tree implementations of DriverNetworkManager.CreateNetwork
have been modified to account for the function signature change.
In updating for the change, the enhancement of adding hostnames to
network namespaces has also been added to the Docker driver, whilst
the default Linux manager does not current implement it.
Speed up client startup, by retrying more until the servers are known.
Currently, if client fingerprinting is fast and finishes before the
client connect to a server, node registration may be delayed by 15
seconds or so!
Ideally, we'd wait until the client discovers the servers and then retry
immediately, but that requires significant code changes.
Here, we simply retry the node registration request every second. That's
basically the equivalent of check if the client discovered servers every
second. Should be a cheap operation.
When testing this change on my local computer and where both servers and
clients are co-located, the time from startup till node registration
dropped from 34 seconds to 8 seconds!
When creating a TCP proxy bridge for Connect tasks, we are at the
mercy of either end for managing the connection state. For long
lived gRPC connections the proxy could reasonably expect to stay
open until the context was cancelled. For the HTTP connections used
by connect native tasks, we experience connection disconnects.
The proxy gets recreated as needed on follow up requests, however
we also emit a WARN log when the connection is broken. This PR
lowers the WARN to a TRACE, because these disconnects are to be
expected.
Ideally we would be able to proxy at the HTTP layer, however Consul
or the connect native task could be configured to expect mTLS, preventing
Nomad from MiTM the requests.
We also can't mange the proxy lifecycle more intelligently, because
we have no control over the HTTP client or server and how they wish
to manage connection state.
What we have now works, it's just noisy.
Fixes#10933
This PR implements a new "System Batch" scheduler type. Jobs can
make use of this new scheduler by setting their type to 'sysbatch'.
Like the name implies, sysbatch can be thought of as a hybrid between
system and batch jobs - it is for running short lived jobs intended to
run on every compatible node in the cluster.
As with batch jobs, sysbatch jobs can also be periodic and/or parameterized
dispatch jobs. A sysbatch job is considered complete when it has been run
on all compatible nodes until reaching a terminal state (success or failed
on retries).
Feasibility and preemption are governed the same as with system jobs. In
this PR, the update stanza is not yet supported. The update stanza is sill
limited in functionality for the underlying system scheduler, and is
not useful yet for sysbatch jobs. Further work in #4740 will improve
support for the update stanza and deployments.
Closes#2527
When the client launches, use a consistent read to fetch its own allocs,
but allow stale read afterwards as long as reads don't revert into older
state.
This change addresses an edge case affecting restarting client. When a
client restarts, it may fetch a stale data concerning its allocs: allocs
that have completed prior to the client shutdown may still have "run/running"
desired/client status, and have the client attempt to re-run again.
An alternative approach is to track the indices such that the client
set MinQueryIndex on the maximum index the client ever saw, or compare
received allocs against locally restored client state. Garbage
collection complicates this approach (local knowledge is not complete),
and the approach still risks starting "dead" allocations (e.g. the
allocation may have been placed when client just restarted and have
already been reschuled by the time the client started. This approach
here is effective against all kinds of stalness problems with small
overhead.
This PR fixes a bug where the underlying Envoy process of a Connect gateway
would consume a full core of CPU if there is more than one sidecar or gateway
in a group. The utilization was being caused by Consul injecting an envoy_ready_listener
on 127.0.0.1:8443, of which only one of the Envoys would be able to bind to.
The others would spin in a hot loop trying to bind the listener.
As a workaround, we now specify -address during the Envoy bootstrap config
step, which is how Consul maps this ready listener. Because there is already
the envoy_admin_listener, and we need to continue supporting running gateways
in host networking mode, and in those case we want to use the same port
value coming from the service.port field, we now bind the admin listener to
the 127.0.0.2 loop-back interface, and the ready listener takes 127.0.0.1.
This shouldn't make a difference in the 99.999% use case where envoy is
being run in its official docker container. Advanced users can reference
${NOMAD_ENVOY_ADMIN_ADDR_<service>} (as they 'ought to) if needed,
as well as the new variable ${NOMAD_ENVOY_READY_ADDR_<service>} for the
envoy_ready_listener.
Adds missing interpolation step to the `meta` blocks when building the task
environment. Also fixes incorrect parameter order in the test assertion and
adds diagnostics to the test.
There are bits of logic in callers of RemoveWorkload on group/task
cleanup hooks which call RemoveWorkload with the "Canary" version
of the workload, in case the alloc is marked as a Canary. This logic
triggers an extra sync with Consul, and also doesn't do the intended
behavior - for which no special casing is necessary anyway. When the
workload is marked for removal, all associated services and checks
will be removed regardless of the Canary status, because the service
and check IDs do not incorporate the canary-ness in the first place.
The only place where canary-ness matters is when updating a workload,
where we need to compute the hash of the services and checks to determine
whether they have been modified, the Canary flag of which is a part of
that.
Fixes#10842
recover
This code just seems incorrect. As it stands today it reports a
successful restore if RecoverTask fails and then DestroyTask succeeds.
This can result in a really annoying bug where it then calls RecoverTask
again, whereby it will probably get ErrTaskNotFound and call DestroyTask
once more.
I think the only reason this has not been noticed so far is because most
drivers like Docker will return Success, then nomad will call
RecoverTask, get an error (not found) and call DestroyTask again, and
get a ErrTasksNotFound err.
This PR makes it so that Nomad will automatically set the CONSUL_TLS_SERVER_NAME
environment variable for Connect native tasks running in bridge networking mode
where Consul has TLS enabled. Because of the use of a unix domain socket for
communicating with Consul when in bridge networking mode, the server name is
a file name instead of something compatible with the mTLS certificate Consul
will authenticate against. "localhost" is by default a compatible name, so Nomad
will set the environment variable to that.
Fixes#10804
Running the `client/allocrunner` tests fail to compile on macOS because the
CNI test file depends on the CNI network configurator, which is in a
Linux-only file.
When `network.mode = "bridge"`, we create a pause container in Docker with no
networking so that we have a process to hold the network namespace we create
in Nomad. The default `/etc/hosts` file of that pause container is then used
for all the Docker tasks that share that network namespace. Some applications
rely on this file being populated.
This changeset generates a `/etc/hosts` file and bind-mounts it to the
container when Nomad owns the network, so that the container's hostname has an
IP in the file as expected. The hosts file will include the entries added by
the Docker driver's `extra_hosts` field.
In this changeset, only the Docker task driver will take advantage of this
option, as the `exec`/`java` drivers currently copy the host's `/etc/hosts`
file and this can't be changed without breaking backwards compatibility. But
the fields are available in the task driver protobuf for community task
drivers to use if they'd like.
This updates `client.Ready()` so it returns once the client node got
registered at the servers. Previously, it returns when the
fingerprinters first batch completes, wtihout ensuring that the node is
stored in the Raft data. The tests may fail later when it with unknown
node errors later.
`client.Reedy()` seem to be only called in CSI and some client stats
now.
This class of bug, assuming client is registered without checking, is a
source of flakiness elsewhere. Other tests use other mechanisms for
checking node readiness, though not consistently.
This PR changes Nomad's wrapper around the Consul NamespaceAPI so that
it will detect if the Consul Namespaces feature is enabled before making
a request to the Namespaces API. Namespaces are not enabled in Consul OSS,
and require a suitable license to be used with Consul ENT.
Previously Nomad would check for a 404 status code when makeing a request
to the Namespaces API to "detect" if Consul OSS was being used. This does
not work for Consul ENT with Namespaces disabled, which returns a 500.
Now we avoid requesting the namespace API altogether if Consul is detected
to be the OSS sku, or if the Namespaces feature is not licensed. Since
Consul can be upgraded from OSS to ENT, or a new license applied, we cache
the value for 1 minute, refreshing on demand if expired.
Fixes https://github.com/hashicorp/nomad-enterprise/issues/575
Note that the ticket originally describes using attributes from https://github.com/hashicorp/nomad/issues/10688.
This turns out not to be possible due to a chicken-egg situation between
bootstrapping the agent and setting up the consul client. Also fun: the
Consul fingerprinter creates its own Consul client, because there is no
[currently] no way to pass the agent's client through the fingerprint factory.
Track usage of incoming streams on a connection. Connections without
reference counts get marked as unused and reaped in a periodic job.
This fixes a bug where `alloc exec` and `alloc fs` sessions get terminated
unexpectedly. Previously, when a client heartbeats switches between
servers, the pool connection reaper eventually identifies the connection
as unused and closes it even if it has an active exec/fs sessions.
Fixes#10579
This PR implements first-class support for Nomad running Consul
Connect Mesh Gateways. Mesh gateways enable services in the Connect
mesh to make cross-DC connections via gateways, where each datacenter
may not have full node interconnectivity.
Consul docs with more information:
https://www.consul.io/docs/connect/gateways/mesh-gateway
The following group level service block can be used to establish
a Connect mesh gateway.
service {
connect {
gateway {
mesh {
// no configuration
}
}
}
}
Services can make use of a mesh gateway by configuring so in their
upstream blocks, e.g.
service {
connect {
sidecar_service {
proxy {
upstreams {
destination_name = "<service>"
local_bind_port = <port>
datacenter = "<datacenter>"
mesh_gateway {
mode = "<mode>"
}
}
}
}
}
}
Typical use of a mesh gateway is to create a bridge between datacenters.
A mesh gateway should then be configured with a service port that is
mapped from a host_network configured on a WAN interface in Nomad agent
config, e.g.
client {
host_network "public" {
interface = "eth1"
}
}
Create a port mapping in the group.network block for use by the mesh
gateway service from the public host_network, e.g.
network {
mode = "bridge"
port "mesh_wan" {
host_network = "public"
}
}
Use this port label for the service.port of the mesh gateway, e.g.
service {
name = "mesh-gateway"
port = "mesh_wan"
connect {
gateway {
mesh {}
}
}
}
Currently Envoy is the only supported gateway implementation in Consul.
By default Nomad client will run the latest official Envoy docker image
supported by the local Consul agent. The Envoy task can be customized
by setting `meta.connect.gateway_image` in agent config or by setting
the `connect.sidecar_task` block.
Gateways require Consul 1.8.0+, enforced by the Nomad scheduler.
Closes#9446
When `nomad volume create` was introduced in Nomad 1.1.0, we changed the
volume spec to take a list of capabilities rather than a single capability, to
meet the requirements of the CSI spec. When a volume is registered via `nomad
volume register`, we should be using the same fields to validate the volume
with the controller plugin.
This PR adds new probes for detecting these new Consul related attributes:
Consul namespaces are a Consul enterprise feature that may be disabled depending
on the enterprise license associated with the Consul servers. Having this attribute
available will enable Nomad to properly decide whether to query the Consul Namespace
API.
Consul connect must be explicitly enabled before Connect APIs will work. Currently
Nomad only checks for a minimum Consul version. Having this attribute available will
enable Nomad to properly schedule Connect tasks only on nodes with a Consul agent that
has Connect enabled.
Consul connect requires the grpc port to be explicitly set before Connect APIs will work.
Currently Nomad only checks for a minimal Consul version. Having this attribute available
will enable Nomad to schedule Connect tasks only on nodes with a Consul agent that has
the grpc listener enabled.
This PR refactors the ConsulFingerprint implementation, breaking individual attributes
into individual functions to make testing them easier. This is in preparation for
additional extractors about to be added. Behavior should be otherwise unchanged.
It adds the attribute consul.sku, which can be used to differentiate between Consul
OSS vs Consul ENT.
Include the VolumeCapability.MountVolume data in
ControllerPublishVolume, CreateVolume, and ValidateVolumeCapabilities
RPCs sent to the CSI controller. The previous behavior was to only
include the MountVolume capability in the NodeStageVolume request, which
on some CSI implementations would be rejected since the Volume was not
originally provisioned with the specific mount capabilities requested.
Follow up to memory oversubscription - expose an env-var to indicate when memory oversubscription is enabled and what the limit is.
This will be helpful for setting hints to app for memory management.
Co-authored-by: Seth Hoenig <shoenig@hashicorp.com>
This commit ensures Nomad captures the task code more reliably even when the task is killed. This issue affect to `raw_exec` driver, as noted in https://github.com/hashicorp/nomad/issues/10430 .
We fix this issue by ensuring that the TaskRunner only calls `driver.WaitTask` once. The TaskRunner monitors the completion of the task by calling `driver.WaitTask` which should return the task exit code on completion. However, it also could return a "context canceled" error if the agent/executor is shutdown.
Previously, when a task is to be stopped, the killTask path makes two WaitTask calls, and the second returns "context canceled" occasionally because of a "race" in task shutting down and depending on driver, and how fast it shuts down after task completes.
By having a single WaitTask call and consistently waiting for the task, we ensure we capture the exit code reliably before the executor is shutdown or the contexts expired.
I opted to change the TaskRunner implementation to avoid changing the driver interface or requiring 3rd party drivers to update.
Additionally, the PR ensures that attempts to kill the task terminate when the task "naturally" dies. Without this change, if the task dies at the right moment, the `killTask` call may retry to kill an already-dead task for up to 5 minutes before giving up.
Add a new driver capability: RemoteTasks.
When a task is run by a driver with RemoteTasks set, its TaskHandle will
be propagated to the server in its allocation's TaskState. If the task
is replaced due to a down node or draining, its TaskHandle will be
propagated to its replacement allocation.
This allows tasks to be scheduled in remote systems whose lifecycles are
disconnected from the Nomad node's lifecycle.
See https://github.com/hashicorp/nomad-driver-ecs for an example ECS
remote task driver.
This PR wraps the use of the consul envoy bootstrap command in
an expoenential backoff closure, configured to timeout after 60
seconds. This is an increase over the current behavior of making
3 attempts over 6 seconds.
Should help with #10451
Similar to a bugfix made for the services hook, we need to always
set the script checks hook, in case a task is initially launched
without script checks, but then updated to include script checks.
The scipt checks hook is the thing that handles that new registration.
(cherry-picked from ent without _ent things)
This is part 2/4 of e2e tests for Consul Namespaces. Took a
first pass at what the parameterized tests can look like, but
only on the ENT side for this PR. Will continue to refactor
in the next PRs.
Also fixes 2 bugs:
- Config Entries registered by Nomad Server on job registration
were not getting Namespace set
- Group level script checks were not getting Namespace set
Those changes will need to be copied back to Nomad OSS.
Nomad OSS + no ACLs (previously, needs refactor)
Nomad ENT + no ACLs (this)
Nomad OSS + ACLs (todo)
Nomad ENT + ALCs (todo)
This PR introduces the /v1/search/fuzzy API endpoint, used for fuzzy
searching objects in Nomad. The fuzzy search endpoint routes requests
to the Nomad Server leader, which implements the Search.FuzzySearch RPC
method.
Requests to the fuzzy search API are based on the api.FuzzySearchRequest
object, e.g.
{
"Text": "ed",
"Context": "all"
}
Responses from the fuzzy search API are based on the api.FuzzySearchResponse
object, e.g.
{
"Index": 27,
"KnownLeader": true,
"LastContact": 0,
"Matches": {
"tasks": [
{
"ID": "redis",
"Scope": [
"default",
"example",
"cache"
]
}
],
"evals": [],
"deployment": [],
"volumes": [],
"scaling_policy": [],
"images": [
{
"ID": "redis:3.2",
"Scope": [
"default",
"example",
"cache",
"redis"
]
}
]
},
"Truncations": {
"volumes": false,
"scaling_policy": false,
"evals": false,
"deployment": false
}
}
The API is tunable using the new server.search stanza, e.g.
server {
search {
fuzzy_enabled = true
limit_query = 200
limit_results = 1000
min_term_length = 5
}
}
These values can be increased or decreased, so as to provide more
search results or to reduce load on the Nomad Server. The fuzzy search
API can be disabled entirely by setting `fuzzy_enabled` to `false`.
on Linux systems this is derived from the configure cpuset cgroup parent (defaults to /nomad)
for non Linux systems and Linux systems where cgroups are not enabled, the client defaults to using all cores
The goal is to always find an interface with an address, preferring
sandbox interfaces, but falling back to the first address found.
A test was added against a known CNI plugin output that was not handled
correctly before.
Registration of Nomad volumes previously allowed for a single volume
capability (access mode + attachment mode pair). The recent `volume create`
command requires that we pass a list of requested capabilities, but the
existing workflow for claiming volumes and attaching them on the client
assumed that the volume's single capability was correct and unchanging.
Add `AccessMode` and `AttachmentMode` to `CSIVolumeClaim`, use these fields to
set the initial claim value, and add backwards compatibility logic to handle
the existing volumes that already have claims without these fields.
This PR adds the common OSS changes for adding support for Consul Namespaces,
which is going to be a Nomad Enterprise feature. There is no new functionality
provided by this changeset and hopefully no new bugs.
In order to support new node RPCs, we need to fingerprint plugin capabilities
in more detail. This changeset mirrors recent work to fingerprint controller
capabilities, but is not yet in use by any Nomad RPC.
In order to support new controller RPCs, we need to fingerprint volume
capabilities in more detail and perform controller RPCs only when the specific
capability is present. This fixes a bug in Ceph support where the plugin can
only suport create/delete but we assume that it also supports attach/detach.
This commit includes a new test client that allows overriding the RPC
protocols. Only the RPCs that are passed in are registered, which lets you
implement a mock RPC in the server tests. This commit includes an example of
this for the ClientCSI RPC server.
Use the MemoryMaxMB as the LinuxResources limit. This is intended to ease
drivers implementation and adoption of the features: drivers that use
`resources.LinuxResources.MemoryLimitBytes` don't need to be updated.
Drivers that use NomadResources will need to updated to track the new
field value. Given that tasks aren't guaranteed to use up the excess
memory limit, this is a reasonable compromise.
Add a `PerAlloc` field to volume requests that directs the scheduler to test
feasibility for volumes with a source ID that includes the allocation index
suffix (ex. `[0]`), rather than the exact source ID.
Read the `PerAlloc` field when making the volume claim at the client to
determine if the allocation index suffix (ex. `[0]`) should be added to the
volume source ID.
Allow for readiness type checks by configuring nomad to ignore warnings
or errors reported by a service check. This allows the deployment to
progress and while Consul handles introducing the sercive into a
resource pool once the check passes.
This PR implements Nomad built-in support for running Consul Connect
terminating gateways. Such a gateway can be used by services running
inside the service mesh to access "legacy" services running outside
the service mesh while still making use of Consul's service identity
based networking and ACL policies.
https://www.consul.io/docs/connect/gateways/terminating-gateway
These gateways are declared as part of a task group level service
definition within the connect stanza.
service {
connect {
gateway {
proxy {
// envoy proxy configuration
}
terminating {
// terminating-gateway configuration entry
}
}
}
}
Currently Envoy is the only supported gateway implementation in
Consul. The gateay task can be customized by configuring the
connect.sidecar_task block.
When the gateway.terminating field is set, Nomad will write/update
the Configuration Entry into Consul on job submission. Because CEs
are global in scope and there may be more than one Nomad cluster
communicating with Consul, there is an assumption that any terminating
gateway defined in Nomad for a particular service will be the same
among Nomad clusters.
Gateways require Consul 1.8.0+, checked by a node constraint.
Closes#9445
Most allocation hooks don't need to know when a single task within the
allocation is restarted. The check watcher for group services triggers the
alloc runner to restart all tasks, but the alloc runner's `Restart` method
doesn't trigger any of the alloc hooks, including the group service hook. The
result is that after the first time a check triggers a restart, we'll never
restart the tasks of an allocation again.
This commit adds a `RunnerTaskRestartHook` interface so that alloc runner
hooks can act if a task within the alloc is restarted. The only implementation
is in the group service hook, which will force a re-registration of the
alloc's services and fix check restarts.
Connect ingress gateway services were being registered into Consul without
an explicit deterministic service ID. Consul would generate one automatically,
but then Nomad would have no way to register a second gateway on the same agent
as it would not supply 'proxy-id' during envoy bootstrap.
Set the ServiceID for gateways, and supply 'proxy-id' when doing envoy bootstrap.
Fixes#9834
* Throw away result of multierror.Append
When given a *multierror.Error, it is mutated, therefore the return
value is not needed.
* Simplify MergeMultierrorWarnings, use StringBuilder
* Hash.Write() never returns an error
* Remove error that was always nil
* Remove error from Resources.Add signature
When this was originally written it could return an error, but that was
refactored away, and callers of it as of today never handle the error.
* Throw away results of io.Copy during Bridge
* Handle errors when computing node class in test
In 492d62d we prevented poststop tasks from contributing to allocation health
status, which fixed a bug where poststop tasks would prevent a deployment from
ever being marked successful. The patch introduced a regression where prestart
tasks that complete are causing the allocation to be marked unhealthy. This
changeset restores the previous behavior for prestart tasks.
* investigating where to ignore poststop task in alloc health tracker
* ignore poststop when setting latest start time for allocation
* clean up logic
* lifecycle: isolate mocks for poststop deployment test
* lifecycle: update comments in tracker
Co-authored-by: Jasmine Dahilig <jasmine@dahilig.com>
When a client restarts, the network_hook's prerun will call
`CreateNetwork`. Drivers that don't implement their own network manager will
fall back to the default network manager, which doesn't handle the case where
the network namespace is being recreated safely. This results in an error and
the task being restarted for `exec` tasks with `network` blocks (this also
impacts the community `containerd` and probably other community task drivers).
If we get an error when attempting to create the namespace and that error is
because the file already exists and is locked by its process, then we'll
return a `nil` error with the `created` flag set to false, just as we do with
the `docker` driver.
When upgrading from Nomad v0.12.x to v1.0.x, Nomad client will panic on
startup if the node is running Connect enabled jobs. This is caused by
a missing piece of plumbing of the Consul Proxies API interface during the
client restore process.
Fixes#9738
This PR deflakes TestTaskRunner_StatsHook_Periodic tests and adds backoff when the driver closes the channel.
TestTaskRunner_StatsHook_Periodic is currently the most flaky test - failing ~4% of the time (20 out of 486 workflows). A sample failure: https://app.circleci.com/pipelines/github/hashicorp/nomad/14028/workflows/957b674f-cbcc-4228-96d9-1094fdee5b9c/jobs/128563 .
This change has two components:
First, it updates the StatsHook so that it backs off when stats channel is closed. In the context of the test where the mock driver emits a single stats update and closes the channel, the test may make tens of thousands update during the period. In real context, if a driver doesn't implement the stats handler properly or when a task finishes, we may generate way too many Stats queries in a tight loop. Here, the backoff reduces these queries. I've added a failing test that shows 154,458 stats updates within 500ms in https://app.circleci.com/pipelines/github/hashicorp/nomad/14092/workflows/50672445-392d-4661-b19e-e3561ed32746/jobs/129423 .
Second, the test ignores the first stats update after a task exit. Due to the asynchronicity of updates and channel/context use, it's possible that an update is enqueued while the test marks the task as exited, resulting into a spurious update.
Previously, Nomad would optimize out the services task runner
hook for tasks which were initially submitted with no services
defined. This causes a problem when the job is later updated to
include service(s) on that task, which will result in nothing
happening because the hook is not present to handle the service
registration in the .Update.
Instead, always enable the services hook. The group services
alloc runner hook is already always enabled.
Fixes#9707
The client allocation GC API returns a misleading error message when the
allocation exists but is not yet eligible for GC. Make this clear in the error
response.
Note in the docs that the allocation will still show on the server responses.
When a task is restored after a client restart, the template runner will
create a new lease for any dynamic secret (ex. Consul or PKI secrets
engines). But because this lease is being created in the prestart hook, we
don't trigger the `change_mode`.
This changeset uses the the existence of the task handle to detect a
previously running task that's been restored, so that we can trigger the
template `change_mode` if the template is changed, as it will be only with
dynamic secrets.
When we iterate over the interfaces returned from CNI setup, we filter for one
with the `Sandbox` field set. Ensure that if none of the interfaces has that
field set that we still return an available interface.
CNI network configuration is currently only supported on Linux.
For now, add the linux build tag so that the deadcode linter does
not trip over unused CNI stuff on macOS.
Nomad v1.0.0 introduced a regression where the client configurations
for `connect.sidecar_image` and `connect.gateway_image` would be
ignored despite being set. This PR restores that functionality.
There was a missing layer of interpolation that needs to occur for
these parameters. Since Nomad 1.0 now supports dynamic envoy versioning
through the ${NOMAD_envoy_version} psuedo variable, we basically need
to first interpolate
${connect.sidecar_image} => envoyproxy/envoy:v${NOMAD_envoy_version}
then use Consul at runtime to resolve to a real image, e.g.
envoyproxy/envoy:v${NOMAD_envoy_version} => envoyproxy/envoy:v1.16.0
Of course, if the version of Consul is too old to provide an envoy
version preference, we then need to know to fallback to the old
version of envoy that we used before.
envoyproxy/envoy:v${NOMAD_envoy_version} => envoyproxy/envoy:v1.11.2@sha256:a7769160c9c1a55bb8d07a3b71ce5d64f72b1f665f10d81aa1581bc3cf850d09
Beyond that, we also need to continue to support jobs that set the
sidecar task themselves, e.g.
sidecar_task { config { image: "custom/envoy" } }
which itself could include teh pseudo envoy version variable.
Previously, Nomad would fail to startup if the CPU fingerprinter could
not detect the cpu total compute (i.e. cores * mhz). This is common on
some EC2 instance types (graviton class), where the env_aws fingerprinter
will override the detected CPU performance with a more accurate value
anyway.
Instead of crashing on startup, have Nomad use a low default for available
cpu performance of 1000 ticks (e.g. 1 core * 1 GHz). This enables Nomad
to get past the useless cpu fingerprinting on those EC2 instances. The
crashing error message is now a log statement suggesting the setting of
cpu_total_compute in client config.
Fixes#7989
This PR enables job submitters to use interpolation in the connect
block of jobs making use of consul connect. Before, only the name of
the connect service would be interpolated, and only for a few select
identifiers related to the job itself (#6853). Now, all connect fields
can be interpolated using the full spectrum of runtime parameters.
Note that the service name is interpolated at job-submission time,
and cannot make use of values known only at runtime.
Fixes#7221
Previously, every Envoy Connect sidecar would spawn as many worker
threads as logical CPU cores. That is Envoy's default behavior when
`--concurrency` is not explicitly set. Nomad now sets the concurrency
flag to 1, which is sensible for the default cpu = 250 Mhz resources
allocated for sidecar proxies. The concurrency value can be configured
in Client configuration by setting `meta.connect.proxy_concurrency`.
Closes#9341
* upsertaclpolicies
* delete acl policies msgtype
* upsert acl policies msgtype
* delete acl tokens msgtype
* acl bootstrap msgtype
wip unsubscribe on token delete
test that subscriptions are closed after an ACL token has been deleted
Start writing policyupdated test
* update test to use before/after policy
* add SubscribeWithACLCheck to run acl checks on subscribe
* update rpc endpoint to use broker acl check
* Add and use subscriptions.closeSubscriptionFunc
This fixes the issue of not being able to defer unlocking the mutex on
the event broker in the for loop.
handle acl policy updates
* rpc endpoint test for terminating acl change
* add comments
Co-authored-by: Kris Hicks <khicks@hashicorp.com>
Always wait 200ms before calling the Node.UpdateAlloc RPC to send
allocation updates to servers.
Prior to this change we only reset the update ticker when an error was
encountered. This meant the 200ms ticker was running while the RPC was
being performed. If the RPC was slow due to network latency or server
load and took >=200ms, the ticker would tick during the RPC.
Then on the next loop only the select would randomly choose between the
two viable cases: receive an update or fire the RPC again.
If the RPC case won it would immediately loop again due to there being
no updates to send.
When the update chan receive is selected a single update is added to the
slice. The odds are then 50/50 that the subsequent loop will send the
single update instead of receiving any more updates.
This could cause a couple of problems:
1. Since only a small number of updates are sent, the chan buffer may
fill, applying backpressure, and slowing down other client
operations.
2. The small number of updates sent may already be stale and not
represent the current state of the allocation locally.
A risk here is that it's hard to reason about how this will interact
with the 50ms batches on servers when the servers under load.
A further improvement would be to completely remove the alloc update
chan and instead use a mutex to build a map of alloc updates. I wanted
to test the lowest risk possible change on loaded servers first before
making more drastic changes.
While Nomad v0.12.8 fixed `NOMAD_{ALLOC,TASK,SECRETS}_DIR` use in
`template.destination`, interpolating these variables in
`template.source` caused a path escape error.
**Why not apply the destination fix to source?**
The destination fix forces destination to always be relative to the task
directory. This makes sense for the destination as a destination outside
the task directory would be unreachable by the task. There's no reason
to ever render a template outside the task directory. (Using `..` does
allow destinations to escape the task directory if
`template.disable_file_sandbox = true`. That's just awkward and unsafe
enough I hope no one uses it.)
There is a reason to source a template outside a task
directory. At least if there weren't then I can't think of why we
implemented `template.disable_file_sandbox`. So v0.12.8 left the
behavior of `template.source` the more straightforward "Interpolate and
validate."
However, since outside of `raw_exec` every other driver uses absolute
paths for `NOMAD_*_DIR` interpolation, this means those variables are
unusable unless `disable_file_sandbox` is set.
**The Fix**
The variables are now interpolated as relative paths *only for the
purpose of rendering templates.* This is an unfortunate special case,
but reflects the fact that the templates view of the filesystem is
completely different (unconstrainted) vs the task's view (chrooted).
Arguably the values of these variables *should be context-specific.*
I think it's more reasonable to think of the "hack" as templating
running uncontainerized than that giving templates different paths is a
hack.
**TODO**
- [ ] E2E tests
- [ ] Job validation may still be broken and prevent my fix from
working?
**raw_exec**
`raw_exec` is actually broken _a different way_ as exercised by tests in
this commit. I think we should probably remove these tests and fix that
in a followup PR/release, but I wanted to leave them in for the initial
review and discussion. Since non-containerized source paths are broken
anyway, perhaps there's another solution to this entire problem I'm
overlooking?
This PR adds the ability to set HTTP headers when downloading
an artifact from an `http` or `https` resource.
The implementation in `go-getter` is such that a new `HTTPGetter`
must be created for each artifact that sets headers (as opposed
to conveniently setting headers per-request). This PR maintains
the memoization of the default Getter objects, creating new ones
only for artifacts where headers are set.
Closes#9306
The unpublish workflow requires that we know the mode (RW vs RO) if we want to
unpublish the node. Update the hook and the Unpublish RPC so that we mark the
claim for release in a new state but leave the mode alone. This fixes a bug
where RO claims were failing node unpublish.
The core job GC doesn't know the mode, but we don't need it for that workflow,
so add a mode specifically for GC; the volumewatcher uses this as a sentinel
to check whether claims (with their specific RW vs RO modes) need to be claimed.
Even if a plugin sends back an empty `[]*device.DeviceGroup`, it's transformed to `nil` during the RPC. Our custom device plugin is returning empty `FingerprintResponse.Devices` very often. Our temporary fix is to send a dummy `*DeviceGroup` if the slice is empty. This has the effect of never triggering the "first fingerprint" and therefore timing out after 50s.
In turn, this made our node exceed its hearbeat grace period when restarting it, revoking all vault tokens for its allocations, causing a restart of all our allocations because the token couldn't be renewed.
Removing the logic for `f.Devices == nil` does not appear to affect the functionality of the function.
In Nomad v0.12.0, the client added additional fingerprinting around the
presense of the bridge kernel module. The fingerprinter only checked in
`/proc/modules` which is a list of loaded modules. In some cases, the
bridge kernel module is builtin rather than dynamically loaded. The fix
for that case is in #8721. However we were still missing the case where
the bridge module is dynamically loaded, but not yet loaded during the
startup of the Nomad agent. In this case the fingerprinter would believe
the bridge module was unavailable when really it gets loaded on demand.
This PR now has the fingerprinter scan the kernel module dependency file,
which will contain an entry for the bridge module even if it is not yet
loaded.
In summary, the client now looks for the bridge kernel module in
- /proc/modules
- /lib/modules/<kernel>/modules.builtin
- /lib/modules/<kernel>/modules.dep
Closes#8423
Beforehand tasks and field replacements did not have access to the
unique ID of their job or its parent. This adds this information as
new environment variables.
Prior to Nomad 0.12.5, you could use `${NOMAD_SECRETS_DIR}/mysecret.txt` as
the `artifact.destination` and `template.destination` because we would always
append the destination to the task working directory. In the recent security
patch we treated the `destination` absolute path as valid if it didn't escape
the working directory, but this breaks backwards compatibility and
interpolation of `destination` fields.
This changeset partially reverts the behavior so that we always append the
destination, but we also perform the escape check on that new destination
after interpolation so the security hole is closed.
Also, ConsulTemplate test should exercise interpolation
Ensure that the client honors the client configuration for the
`template.disable_file_sandbox` field when validating the jobspec's
`template.source` parameter, and not just with consul-template's own `file`
function.
Prevent interpolated `template.source`, `template.destination`, and
`artifact.destination` fields from escaping file sandbox.
* use msgtype in upsert node
adds message type to signature for upsert node, update tests, remove placeholder method
* UpsertAllocs msg type test setup
* use upsertallocs with msg type in signature
update test usage of delete node
delete placeholder msgtype method
* add msgtype to upsert evals signature, update test call sites with test setup msg type
handle snapshot upsert eval outside of FSM and ignore eval event
remove placeholder upsertevalsmsgtype
handle job plan rpc and prevent event creation for plan
msgtype cleanup upsertnodeevents
updatenodedrain msgtype
msg type 0 is a node registration event, so set the default to the ignore type
* fix named import
* fix signature ordering on upsertnode to match
* consul: advertise cni and multi host interface addresses
* structs: add service/check address_mode validation
* ar/groupservices: fetch networkstatus at hook runtime
* ar/groupservice: nil check network status getter before calling
* consul: comment network status can be nil
As newer versions of Consul are released, the minimum version of Envoy
it supports as a sidecar proxy also gets bumped. Starting with the upcoming
Consul v1.9.X series, Envoy v1.11.X will no longer be supported. Current
versions of Nomad hardcode a version of Envoy v1.11.2 to be used as the
default implementation of Connect sidecar proxy.
This PR introduces a change such that each Nomad Client will query its
local Consul for a list of Envoy proxies that it supports (https://github.com/hashicorp/consul/pull/8545)
and then launch the Connect sidecar proxy task using the latest supported version
of Envoy. If the `SupportedProxies` API component is not available from
Consul, Nomad will fallback to the old version of Envoy supported by old
versions of Consul.
Setting the meta configuration option `meta.connect.sidecar_image` or
setting the `connect.sidecar_task` stanza will take precedence as is
the current behavior for sidecar proxies.
Setting the meta configuration option `meta.connect.gateway_image`
will take precedence as is the current behavior for connect gateways.
`meta.connect.sidecar_image` and `meta.connect.gateway_image` may make
use of the special `${NOMAD_envoy_version}` variable interpolation, which
resolves to the newest version of Envoy supported by the Consul agent.
Addresses #8585#7665
Previously, Nomad was using a hand-made lookup table for looking
up EC2 CPU performance characteristics (core count + speed = ticks).
This data was incomplete and incorrect depending on region. The AWS
API has the correct data but requires API keys to use (i.e. should not
be queried directly from Nomad).
This change introduces a lookup table generated by a small command line
tool in Nomad's tools module which uses the Amazon AWS API.
Running the tool requires AWS_* environment variables set.
$ # in nomad/tools/cpuinfo
$ go run .
Going forward, Nomad can incorporate regeneration of the lookup table
somewhere in the CI pipeline so that we remain up-to-date on the latest
offerings from EC2.
Fixes#7830
Host with systemd-resolved have `/etc/resolv.conf` is a symlink
to `/run/systemd/resolve/stub-resolv.conf`. By bind-mounting
/etc/resolv.conf only, the exec container DNS resolution fail very badly.
This change fixes DNS resolution by binding /run/systemd/resolve as
well.
Note that this assumes that the systemd resolver (default to 127.0.0.53) is
accessible within the container. This is the case here because exec
containers share the same network namespace by default.
Jobs with custom network dns configurations are not affected, and Nomad
will continue to use the job dns settings rather than host one.
When defining a script-check in a group-level service, Nomad needs to
know which task is associated with the check so that it can use the
correct task driver to execute the check.
This PR fixes two bugs:
1) validate service.task or service.check.task is configured
2) make service.check.task inherit service.task if it is itself unset
Fixes#8952
- We previously added these to the client host metrics, but it's useful to have them on all client metrics.
- e.g. so you can exclude draining nodes from charts showing your fleet size.
The current implementation measures RPC request timeout only against
config.RPCHoldTimeout, which is fine for non-blocking requests but will
almost surely be exceeded by long-poll requests that block for minutes
at a time.
This adds an HasTimedOut method on the RPCInfo interface that takes into
account whether the request is blocking, its maximum wait time, and the
RPCHoldTimeout.
This PR adds initial support for running Consul Connect Ingress Gateways (CIGs) in Nomad. These gateways are declared as part of a task group level service definition within the connect stanza.
```hcl
service {
connect {
gateway {
proxy {
// envoy proxy configuration
}
ingress {
// ingress-gateway configuration entry
}
}
}
}
```
A gateway can be run in `bridge` or `host` networking mode, with the caveat that host networking necessitates manually specifying the Envoy admin listener (which cannot be disabled) via the service port value.
Currently Envoy is the only supported gateway implementation in Consul, and Nomad only supports running Envoy as a gateway using the docker driver.
Aims to address #8294 and tangentially #8647
* docker: support group allocated ports
* docker: add new ports driver config to specify which group ports are mapped
* docker: update port mapping docs
When the client-side actions of a CSI client RPC succeed but we get
disconnected during the RPC or we fail to checkpoint the claim state, we want
to be able to retry the client RPC without getting blocked by the client-side
state (ex. mount points) already having been cleaned up in previous calls.
Add a Postrun hook to send the `CSIVolume.Unpublish` RPC to the server. This
may forward client RPCs to the node plugins or to the controller plugins,
depending on whether other allocations on this node have claims on this
volume.
By making clients responsible for running the `CSIVolume.Unpublish` RPC (and
making the RPC available to a `nomad volume detach` command), the
volumewatcher becomes only used by the core GC job and we no longer need
async volume GC from job deregister and node update.
Before, Connect Native Tasks needed one of these to work:
- To be run in host networking mode
- To have the Consul agent configured to listen to a unix socket
- To have the Consul agent configured to listen to a public interface
None of these are a great experience, though running in host networking is
still the best solution for non-Linux hosts. This PR establishes a connection
proxy between the Consul HTTP listener and a unix socket inside the alloc fs,
bypassing the network namespace for any Connect Native task. Similar to and
re-uses a bunch of code from the gRPC listener version for envoy sidecar proxies.
Proxy is established only if the alloc is configured for bridge networking and
there is at least one Connect Native task in the Task Group.
Fixes#8290
adds in oss components to support enterprise multi-vault namespace feature
upgrade specific doc on vault multi-namespaces
vault docs
update test to reflect new error
The NodePublish workflow currently creates the target path and its parent
directory. However, the CSI specification says that the CO shall ensure the
parent directory of the target path exists, and that the SP shall place the
block device or mounted directory at the target path. Much of our testing has
been with CSI plugins that are more forgiving, but our behavior breaks
spec-compliant CSI plugins.
This changeset ensures we only create the parent directory.
Also fixed the same typo in a test. Fixing the typo fixes the link, but
the link was still broken when running the website locally due to the
trailing slash. It would have worked in prod thanks to redirects, but
using the canonical URL seems ideal.
* ar: support opting into binding host ports to default network IP
* fix config plumbing
* plumb node address into network resource
* struct: only handle network resource upgrade path once
* command/agent/host: collect host data, multi platform
* nomad/structs/structs: new HostDataRequest/Response
* client/agent_endpoint: add RPC endpoint
* command/agent/agent_endpoint: add Host
* api/agent: add the Host endpoint
* nomad/client_agent_endpoint: add Agent Host with forwarding
* nomad/client_agent_endpoint: use findClientConn
This changes forwardMonitorClient and forwardProfileClient to use
findClientConn, which was cribbed from the common parts of those
funcs.
* command/debug: call agent hosts
* command/agent/host: eliminate calling external programs
This fixes a bug where a batch allocation fails to complete if it has
sidecars.
If the only remaining running tasks in an allocations are sidecars - we
must kill them and mark the allocation as complete.
This PR adds the capability of running Connect Native Tasks on Nomad,
particularly when TLS and ACLs are enabled on Consul.
The `connect` stanza now includes a `native` parameter, which can be
set to the name of task that backs the Connect Native Consul service.
There is a new Client configuration parameter for the `consul` stanza
called `share_ssl`. Like `allow_unauthenticated` the default value is
true, but recommended to be disabled in production environments. When
enabled, the Nomad Client's Consul TLS information is shared with
Connect Native tasks through the normal Consul environment variables.
This does NOT include auth or token information.
If Consul ACLs are enabled, Service Identity Tokens are automatically
and injected into the Connect Native task through the CONSUL_HTTP_TOKEN
environment variable.
Any of the automatically set environment variables can be overridden by
the Connect Native task using the `env` stanza.
Fixes#6083
In #7957 we added support for passing a volume context to the controller RPCs.
This is an opaque map that's created by `CreateVolume` or, in Nomad's case,
in the volume registration spec.
However, we missed passing this field to the `NodeStage` and `NodePublish` RPC,
which prevents certain plugins (such as MooseFS) from making node RPCs.
