The fresh deployment of the Redis job took around 20s which is
also the default context timeout on the e2e util that monitors and
waits for a deployment to complete.
The tight timing meant the test often timed out but sometimes
would complete successfully. Increasing the timeout for this
deployment will remove the flakiness.
While working on #25726, I explored a hypothesis that the problem could be
in the state store, but this proved to be a dead end. While I was in this area
of the code I migrated the tests to `shoenig/test`.
Ref: https://github.com/hashicorp/nomad/pull/25726
* Preserve core resources during inplace service alloc updates
When an alloc is running with the core resources specified, and the
alloc is able to be updated in place, the cores it is running on should
be preserved.
This fixes a bug where the allocation's task's core resources
(CPU.ReservedCores) would be recomputed each time the reconciler checked
that the allocation could continue to run on the given node. Under
circumstances where a different core on the node became available before
this check was made, the selection process could compute this new core
as the core to run on, regardless of core the allocation was already
running on. The check takes into account other allocations running on
the node with reserved cores, but cannot check itself.
When this would happen for multiple allocations being evaluated in a
single plan, the selection process would see the other cores being
previously reserved but be unaware of the one it ran on, resulting in
the same core being chosen over and over for each allocation that was
being checked, and updated in the state store (but not on the node).
Once those cores were chosen and committed for multiple allocs, the node
appears to be exhausted on the cores dimension, and it would prevent any
additional allocations from being started on the node.
The reconciler check/computation for allocations that are being updated
in place and have resources.cores defined is effectively a check that
the node has the available cores to run on, not a computation that
should be changed. The fix still performs the check, but once it is
successful any existing ReservedCores are preserved. Because any changes
to this resource is considered a "destructive change", this can be
confidently preserved during the inplace update.
* Adjust reservedCores scheduler test
* Add changelog entry
ResolveToken RPC endpoint was only used by the /acl/token/self API. We should migrate to the WI-aware WhoAmI instead.
---------
Co-authored-by: Tim Gross <tgross@hashicorp.com>
As of April 1, Docker Hub rate limits tightened. With only 10 pulls/hr/IP, we're
likely to encounter test failures. Switch all Docker images getting pulled from
this repository to use the HashiCorp managed registry mirror.
Note that most of our tests in `drivers/docker` don't pull from the remote
registry but load a local image, while others will need to pull from the remote
and fetch different images depending on OS/arch. Refactor the definition of test
task configuration to make it clear which is which, and de-factor some false
sharing of setup functions.
Updates the E2E tests to use that registry by configuring the Docker
daemon. This required changing out a few container images that we don't have in
the registry, but these new images are all smaller. There are a couple of tests
that still use explicitly-tagged `docker.io` images or other third-party
registries, which have been left in place.
Ref: https://hashicorp.atlassian.net/browse/NET-12233
update E2E images to those in the registry mirror
fix windows and docklog test build
fix stopsignal test
mop-up
more mop-up
In #12319 we fixed a bug where updates to the reschedule tracker would be
dropped if the follow-up allocation failed to be placed by the scheduler in the
later evaluation. We did this by mutating the previous allocation's reschedule
tracker. But we did this without copying the previous allocation first and then
making sure the updated copy was in the plan. This is unfortunately unsafe and
corrupts the state store on the server where the scheduler ran; it may cause a
race condition in RPC handlers and it causes the server to be out of sync with
the other servers. This was discovered while trying to make all our tests
race-free, but likely impacts production users.
Copy the previous allocation before updating the reschedule tracker, and swap
out the updated allocation in the plan. This also requires that we include the
reschedule tracker in the "normalized" (stripped-down) allocations we send to
the leader as part of a plan.
Ref: https://github.com/hashicorp/nomad/pull/12319
Fixes: https://hashicorp.atlassian.net/browse/NET-12357
Nomad is released as a Linux arm64 binary, so having a compilation
step on this OS/ARCH within our core test workflow will help
ensure basic arm64 problems do not get into our release branches.
