Repository navigation
test(two-node): pending F18/F19 fixes as simulator node capabilities - #1490
Merged
Merged
Conversation
…in the simulator The simulator's node model gains capabilities for SPDK/ultra changes that exist as PRs but are not in the node image main deploys: - ana_restore_all_groups (F18, SPDK #80): the unfence restores every ANA group the fence changed; - grant_stops_redirect (F19, SPDK #81 + ultra #215): every solo grant releases the LVS, and the release stops a non-leader's hublvol redirect to the lost peer, so the next IO takes the LVS over at once. DEPLOYED_CAPS is empty, matching what is merged, so main's CI stays truthful: F18 and F19 remain known failures for their signature. test_pending_fix_passes runs each with its fix's capability and requires a pass; test_bed_scenario_with_pending_caps runs every bed scenario with all pending capabilities (the stack under test on the bed in c8). When a fix merges into the deployed image, add its capability to DEPLOYED_CAPS and drop the scenario from KNOWN_FAILING and PENDING_FIX. Local (WSL, real FDB): the 11 bed scenarios pass with and without the pending capabilities; F18/F19 pass with them and xfail without. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Contributor
Author
|
On hold, not merging: bed run c8 shows the simulator's client model is optimistic. The real host detects a frozen path per controller via keep-alive (KATO..2xKATO, 7.1 s measured) and the takeover's JC reader can block on the frozen peer's JM until the 8 s NVMe IO timeout; the sim times out each IO KATO after issue and takes over immediately, so it calls F19 fixed while the bed shows a 9 s stall. The model needs these two behaviours before this PR is truthful. Also: the Linter job fails on an unrelated RUF012 in simplyblock_core/utils/ttl_cache.py. |
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
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
2 participants
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.
Simulator: pending F18/F19 fixes as node capabilities
Test code only.
The simulator's node model gains capabilities for SPDK/ultra changes that exist as PRs but are not yet in the node image
maindeploys:ana_restore_all_groupsgrant_stops_redirectsimnode.DEPLOYED_CAPSis empty, matching what is merged, somain's CI stays truthful: F18 and F19 remain known failures, checked against their failure signature.New tests:
test_pending_fix_passes: F18 and F19, each with its fix's capability, must pass.test_bed_scenario_with_pending_caps: every bed scenario with all pending capabilities. This is the stack under test on the bed in c8.When a fix merges into the deployed node image, add its capability to
DEPLOYED_CAPSand drop the scenario fromKNOWN_FAILINGandPENDING_FIX(documented inAGENTS.md).Local results (WSL, real FDB): the 11 bed scenarios pass with and without the pending capabilities; F18 and F19 pass with them and xfail without.
🤖 Generated with Claude Code