zotac/graylog-deploy
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

Fix startup race: wait for the REST API before the first authenticated call

Browse source 
Docker's healthcheck can report graylog-server "healthy" a few seconds
before the REST API is actually ready to serve authenticated requests -
confirmed live: the first gcurl call (step_index_retention) intermittently
got an empty response body, crashing the downstream `python3 -c
"json.load(sys.stdin)"` with "Expecting value: line 1 column 1".

Adds wait_for_api_ready(), polling the same endpoint step_index_retention
already needs (up to 60s) before proceeding, mirroring the existing
retry-loop pattern already used for docker compose pull.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
This commit is contained in:
byrsapty 2026-07-22 22:45:28 +03:00
parent 71ba82e0f0
commit 63e6ce88d0
1 changed files with 19 additions and 0 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

19
install-graylog.sh
View file

@ -208,6 +208,24 @@ gcurl() {
fi
}
wait_for_api_ready() {
# Docker's healthcheck can report the container "healthy" a few seconds
# before Graylog's REST API is actually ready to serve authenticated
# requests (indices/auth subsystems still initializing) - confirmed live:
# the very first gcurl call (step_index_retention) intermittently got an
# empty response, which then crashed the downstream `python3 -c
# "json.load(sys.stdin)"` with "Expecting value: line 1 column 1".
log "Waiting for the Graylog REST API to accept authenticated requests..."
local waited=0
while true; do
gcurl GET /system/indices/index_sets | python3 -c "import json,sys;json.load(sys.stdin)" 2>/dev/null && break
waited=$((waited + 3))
[ "$waited" -ge 60 ] && die "Graylog REST API did not respond with valid JSON within 60s of the container reporting healthy."
sleep 3
done
ok "Graylog REST API is ready"
}
step_index_retention() {
# Graylog's factory default (as of 7.1) keeps 30-40 days of data across up
# to 20 indices - reasonable in general, but risky on a small disk (this
@ -484,6 +502,7 @@ main() {
# aborts the script - gcurl() is always called from inside command
# substitutions downstream, where `die`'s `exit` would only kill a subshell.
resolve_admin_password
wait_for_api_ready
step_index_retention
step_inputs