freshlybakedca.ke
We're seeing some issues with jobs choking up when we have a long-waiting maintenance service
Instead, we think it might be better to have this run every 30 seconds for at most 30 seconds, thus causing stuck jobs not to cause trouble...
+5
-3
packetmix/systems/teal/wiki.nix
+5
-3
packetmix/systems/teal/wiki.nix
···
243
243
systemd.timers.mediawiki-maintenance = {
244
244
wantedBy = [ "timers.target" ];
245
245
timerConfig = {
246
-
OnUnitActiveSec = "5min";
247
-
OnBootSec = "5min";
246
+
OnUnitActiveSec = "30";
247
+
OnBootSec = "30";
248
248
Persistent = false;
249
249
Unit = "mediawiki-maintenance.service";
250
250
};
···
252
252
253
253
systemd.services.mediawiki-maintenance = {
254
254
script = ''
255
-
${config.services.phpfpm.pools.mediawiki.phpPackage}/bin/php ${config.services.mediawiki.finalPackage}/share/mediawiki/maintenance/run.php runJobs --memory-limit 1G --wait
255
+
${config.services.phpfpm.pools.mediawiki.phpPackage}/bin/php ${config.services.mediawiki.finalPackage}/share/mediawiki/maintenance/run.php runJobs --memory-limit 1G --maxtime 30
256
256
'';
257
257
serviceConfig = {
258
+
Type = "oneshot";
259
+
RemainAfterExit = false;
258
260
User = "mediawiki";
259
261
Group = "nginx";
260
262
PrivateTmp = true;
History
3 rounds
0 comments
a.starrysky.fyi
submitted
#2
1 commit
expand
collapse
fix(pm/wiki): Correct poor maintenance behaviors
We're seeing some issues with jobs choking up when we have a
long-waiting maintenance service
Instead, we think it might be better to have this run every 30 seconds
for at most 30 seconds, thus causing stuck jobs not to cause trouble...
expand 0 comments
pull request successfully merged
a.starrysky.fyi
submitted
#1
1 commit
expand
collapse
fix(pm/wiki): Correct poor maintenance behaviors
We're seeing some issues with jobs choking up when we have a
long-waiting maintenance service
Instead, we think it might be better to have this run every 30 seconds
for at most 30 seconds, thus causing stuck jobs not to cause trouble...
expand 0 comments
a.starrysky.fyi
submitted
#0
1 commit
expand
collapse
fix(pm/wiki): Correct poor maintenance behaviors
We're seeing some issues with jobs choking up when we have a
long-waiting maintenance service
Instead, we think it might be better to have this run every 30 seconds
for at most 30 seconds, thus causing stuck jobs not to cause trouble...