feat(cpp): run C++ SDK tests only against the VSR server - #3844
Merged
Conversation
Codecov Report❌ Patch coverage is Additional details and impacted files@@ Coverage Diff @@
## master #3844 +/- ##
============================================
- Coverage 76.98% 76.09% -0.89%
- Complexity 1021 1316 +295
============================================
Files 1366 1378 +12
Lines 175906 177209 +1303
Branches 145834 146337 +503
============================================
- Hits 135413 134843 -570
- Misses 36619 38358 +1739
- Partials 3874 4008 +134
🚀 New features to boost your workflow:
|
slbotbm
previously approved these changes
Aug 8, 2026
hubcio
force-pushed
the
ci/cpp-vsr-only
branch
from
August 10, 2026 10:06
e36e537 to
1eb24cd
Compare
numinnex
previously approved these changes
Aug 10, 2026
mmodzelewski
previously approved these changes
Aug 10, 2026
spetz
previously approved these changes
Aug 10, 2026
The C++ bindings were the last SDK exercised only against the legacy server. Moving the e2e and BDD lanes to a vsr-built iggy-server-ng meant rewriting every legacy-only expectation: logout drops the client-table entry, flush_unsaved_buffer denies typed, cluster metadata is readable pre-login, group ids are minted monotonically, and a send reports the written offsets. Four gaps surfaced as bugs. Stats hardcoded clients_count to 0 because the sync read path could not run the cross-shard ListClients gather. A poll against a missing stream or topic answered the generic InvalidIdentifier. An over-count partition delete acked a no-op instead of InvalidPartitionsCount. A purge acked before its counters moved, so an immediate read saw pre-purge totals; the reset now runs in the apply, gated on the purge generation because left-right replays it on the second buffer after the ack.
server-ng now denies a poll of a missing stream or topic with the typed not-found rejection instead of a status-0 empty poll, so the TCP suite pin of the old empty-poll reply fails with code 2010. Fold the TCP case into the throw expectation the HTTP variant already carries.
hubcio
dismissed stale reviews from spetz, mmodzelewski, and numinnex
via
August 10, 2026 13:57
758e581
hubcio
force-pushed
the
ci/cpp-vsr-only
branch
from
August 10, 2026 13:57
1eb24cd to
758e581
Compare
numinnex
approved these changes
Aug 10, 2026
spetz
approved these changes
Aug 10, 2026
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
The C++ bindings were the last SDK exercised only against the legacy
server. Moving the e2e and BDD lanes to a vsr-built iggy-server-ng
meant rewriting every legacy-only expectation: logout drops the
client-table entry, flush_unsaved_buffer denies typed, cluster
metadata is readable pre-login, group ids are minted monotonically,
and a send reports the written offsets.
Four gaps surfaced as bugs. Stats hardcoded clients_count to 0 because
the sync read path could not run the cross-shard ListClients gather. A
poll against a missing stream or topic answered the generic
InvalidIdentifier. An over-count partition delete acked a no-op
instead of InvalidPartitionsCount. A purge acked before its counters
moved, so an immediate read saw pre-purge totals; the reset now runs
in the apply, gated on the purge generation because left-right replays
it on the second buffer after the ack.