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.
Nick Ethier