With a suitably slow system -- say, if you're waiting on a Windows
Terminal to output text inside a virtual machine -- the expected order
of things can be thrown off. One specific way this can happen is that
the worker can attempt to dial and have its dial fail because of the
slowness causing the dial context to time out. If, because of the
slowness, we also haven't released the log gate yet, then the event with
the failure information will be queued, along with the context that was
used.
Unfortunately, in some of these error cases, the context that was used
was the dial context instead of the system base context. In many other
places in the function it was the system base context so this is just a
mismatch, probably from code written at different times.
Normally this wouldn't be a problem as we'd fall back to the underlying
logger, but when we release the log gate, things happen differently:
it's a synchronous function and on error it causes us to abandon
system startup entirely.
This commit fixes the issue with using the incorrect context. It's an
open question whether we should change the behavior around errors when
replaying queued events, falling back to the underlying logger instead
of erroring.
* chore: Fix comment typo
* test(e2e): Update to use new credential library type
* test(e2e): Update to use new field name
* test(bats): Remove tests that use deprecated subcommand
* chore: Update to 0.14.0
* test(credentiallibraries): Update to use new credential library type
* doc: Add deprecation notices for 0.14.0
* fixup! doc: Add deprecation notices for 0.14.0
With the KMS changes this should never have worked at all (because the
different databases means the credentials conveyed after the KMS-based
authentication would never be valid). However, it was sometimes passing
because of a race condition: the worker could be found in the database
before the new credentials were returned to the worker, causing the
worker to start at the KMS step with the second controller.
This changes the test to use the same database for two controllers, but
to only ever have one running at a time, and uses a sequence of two
changed (or unchanged, depending) status updates from the worker to
ensure that the reload was successful.
* chore(enos): Move terraform-enos-aws-infra to boundary
The terraform-enos-aws-infra repo was a terraform module to set up the base aws infrastructure. Vault made some local updates that might mitigate flakiness we've been seeing ("can't delete due to dependencies" error). We are replacing our usage of that module Vault's local copy.
* chore(enos): Modify service user to allow DeleteRoute
* fixup! chore(enos): Move terraform-enos-aws-infra to boundary
* chore(enos): Rename variable for clarity
The existing pattern was subject to a bug whereby a lock
value was copied every time the value was passed around.
This was happening because of the use of an embedded
protobuf type. To avoid copying a mutex, use a pointer
type in the embedding.
Also fixes a few linter warnings complaining about the use
of unkeyed literals.
* fix (events): convert multierror to stdlib error
multierror errors will be converted to a stdlib error (because
multierror doesn't support json marshaling).
* refact (errors): convert from multierror to stdlib errors.Join(...)
* refact: convert from pkg/errors to stdlib error wrapping
Jeff Mitchell
Hugo
Michael Li
Irena Rindos
Devon Anderson
Timothy Messier
Jim
irenarindos
Michele Degges
Devon Anderson
April Ayres
daniandDanielle Miu
Damian Debkowski
Dan Heath
Sepehr
April Ayres
Louis Ruch
Johan Brandhorst-Satzkorn
Adam BouhmadandDan Heath