Skip to content

fix(ci): preserve both attempt logs when retrying flaky tests #20439

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
ludamad merged 4 commits into from
Feb 16, 2026
Merged

fix(ci): preserve both attempt logs when retrying flaky tests #20439

Changes from all commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
325f490
fix(ci): give each retry attempt its own unique log key
ludamad Feb 12, 2026
22716ec
Merge branch 'merge-train/spartan' into ad/fix-retry-reporting
ludamad Feb 12, 2026
b7b5c92
Merge branch 'merge-train/spartan' into ad/fix-retry-reporting
ludamad Feb 13, 2026
cbaee3b
Merge branch 'merge-train/spartan' into ad/fix-retry-reporting
ludamad Feb 13, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
32 changes: 29 additions & 3 deletions ci3/run_test_cmd
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -119,6 +119,21 @@ function publish_log_final {
cat $tmp_file 2>/dev/null | cache_persistent $log_key $expire
}

# Finalize the current log and start a fresh one with a new unique key.
function rotate_log {
if [ "$CI_REDIS_AVAILABLE" -eq 1 ]; then
publish_log_final "$@"
fi
log_key=$(uuid)
log_info=" ($(ci_term_link $log_key))"
> $tmp_file
if [ "$live_logging" -eq 1 ]; then
kill ${publish_pid:-} &>/dev/null
live_publish_log &
publish_pid=$!
fi
}

function live_publish_log {
# Not replacing previous trap as we run this function in the background.
trap 'kill $sleep_pid &>/dev/null; exit' SIGTERM SIGINT
Expand Down Expand Up @@ -336,12 +351,23 @@ flake_group_id=$(echo "$test_entries" | jq -r '.flake_group_id // empty' | head
if [ -z "$owners" ]; then
fail
else
echo -e "${yellow}RETRYING${reset}${log_info:-}: $test_cmd"
failure_log_key=$log_key
failure_log_info=$log_info
rotate_log $((60 * 60 * 24 * 7 * 12))

echo -e "${yellow}RETRYING${reset}${log_info}: $test_cmd"

run_test

# Test passed. Signal it as a flake, but pass.
[ $code -eq 0 ] && flake
if [ $code -eq 0 ]; then
# Publish the retry's log, then point back at the failure for flake reporting.
if [ "$CI_REDIS_AVAILABLE" -eq 1 ]; then
publish_log_final
fi
log_key=$failure_log_key
log_info=$failure_log_info
flake
fi

# Otherwise we failed twice in a row, so hard fail.
fail
Expand Down
Loading