The spread iterator can panic when processing an evaluation, resulting
in an unrecoverable state in the cluster. Whenever a panicked server
restarts and quorum is restored, the next server to dequeue the
evaluation will panic.
To trigger this state:
* The job must have `max_parallel = 0` and a `canary >= 1`.
* The job must not have a `spread` block.
* The job must have a previous version.
* The previous version must have a `spread` block and at least one
failed allocation.
In this scenario, the desired changes include `(place 1+) (stop
1+), (ignore n) (canary 1)`. Before the scheduler can place the canary
allocation, it tries to find out which allocations can be
stopped. This passes back through the stack so that we can determine
previous-node penalties, etc. We call `SetJob` on the stack with the
previous version of the job, which will include assessing the `spread`
block (even though the results are unused). The task group spread info
state from that pass through the spread iterator is not reset when we
call `SetJob` again. When the new job version iterates over the
`groupPropertySets`, it will get an empty `spreadAttributeMap`,
resulting in an unexpected nil pointer dereference.
This changeset resets the spread iterator internal state when setting
the job, logging with a bypass around the bug in case we hit similar
cases, and a test that panics the scheduler without the patch.
Add new namespace ACL requirement for the /v1/jobs/parse endpoint and
return early if HCLv2 parsing fails.
The endpoint now requires the new `parse-job` ACL capability or
`submit-job`.
go-getter creates a circular dependency between a Client and Getter,
which means each is inherently thread-unsafe if you try to re-use
on or the other.
This PR fixes Nomad to no longer make use of the default Getter objects
provided by the go-getter package. Nomad must create a new Client object
on every artifact download, as the Client object controls the Src and Dst
among other things. When Caling Client.Get, the Getter modifies its own
Client reference, creating the circular reference and race condition.
We can still achieve most of the desired connection caching behavior by
re-using a shared HTTP client with transport pooling enabled.
When an allocation is updated, the job summary for the associated job
is also updated. CSI uses the job summary to set the expected count
for controller and node plugins. We incorrectly used the allocation's
server status instead of the job status when deciding whether to
update or remove the job from the plugins. This caused a node drain or
other terminal state for an allocation to clear the expected count for
the entire plugin.
Use the job status to guide whether to update or remove the expected
count.
The existing CSI tests for the state store incorrectly modeled the
updates we received from servers vs those we received from clients,
leading to test assertions that passed when they should not.
Rework the tests to clarify each step in the lifecycle and rename CSI state
store functions for clarity
Processing an evaluation is nearly a pure function over the state
snapshot, but we randomly shuffle the nodes. This means that
developers can't take a given state snapshot and pass an evaluation
through it and be guaranteed the same plan results.
But the evaluation ID is already random, so if we use this as the seed
for shuffling the nodes we can greatly reduce the sources of
non-determinism. Unfortunately golang map iteration uses a global
source of randomness and not a goroutine-local one, but arguably
if the scheduler behavior is impacted by this, that's a bug in the
iteration.
If processing a specific evaluation causes the scheduler (and
therefore the entire server) to panic, that evaluation will never
get a chance to be nack'd and cleared from the state store. It will
get dequeued by another scheduler, causing that server to panic, and
so forth until all servers are in a panic loop. This prevents the
operator from intervening to remove the evaluation or update the
state.
Recover the goroutine from the top-level `Process` methods for each
scheduler so that this condition can be detected without panicking the
server process. This will lead to a loop of recovering the scheduler
goroutine until the eval can be removed or nack'd, but that's much
better than taking a downtime.
PR #11956 implemented a new mTLS RPC check to validate the role of the
certificate used in the request, but further testing revealed two flaws:
1. client-only endpoints did not accept server certificates so the
request would fail when forwarded from one server to another.
2. the certificate was being checked after the request was forwarded,
so the check would happen over the server certificate, not the
actual source.
This commit checks for the desired mTLS level, where the client level
accepts both, a server or a client certificate. It also validates the
cercertificate before the request is forwarded.
Non-CSI garbage collection tasks on the server only log the cutoff
index in the case where it's not a forced GC from `nomad system gc`.
Do the same for CSI for consistency.
