conjurer: give librarian 60s termination grace for checkpointing

The librarian now checkpoints an in-progress search on SIGTERM and exits within CONJURER_LIBRARIAN_GRACEFUL_TIMEOUT (default 45s). Raise k8s terminationGracePeriodSeconds to 60 so that graceful checkpoint isn't cut short by SIGKILL - otherwise the long DB scan restarts instead of resuming. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
build: automatic update of conjurer
2026-08-02 22:11:05 +02:00 · 2026-08-02 17:39:36 +00:00 · 2026-08-02 19:36:14 +02:00 · 2026-08-02 17:05:31 +00:00 · 2026-08-02 17:03:28 +00:00 · 2026-08-02 15:33:09 +00:00
2 changed files with 13 additions and 4 deletions
@@ -6,6 +6,6 @@ resources:
  - bot.yaml
 images:
  - name: gitea.czernobog.pl/gitea/conjurer-librarian
-    newTag: 5d321f2f
+    newTag: 40605b95
  - name: gitea.czernobog.pl/gitea/conjurer-bot
-    newTag: 5d321f2f
+    newTag: 8e18071b
@@ -10,6 +10,11 @@ spec:
    metadata: { labels: { app: librarian } }
    spec:
      imagePullSecrets: [{ name: gitea-registry }]
+      # On SIGTERM the librarian checkpoints the running search and exits within
+      # CONJURER_LIBRARIAN_GRACEFUL_TIMEOUT (default 45s). Give k8s enough grace
+      # to let that finish before it SIGKILLs - otherwise a long scan's checkpoint
+      # is cut short and the search restarts from scratch instead of resuming.
+      terminationGracePeriodSeconds: 60
      containers:
        - name: librarian
          image: gitea.czernobog.pl/gitea/conjurer-librarian:e243777e
@@ -33,8 +38,12 @@ spec:
            - { name: doi,   mountPath: /doi, readOnly: true }
            - { name: state, mountPath: /lib_temp_files }
          resources:
-            requests: { cpu: "100m", memory: "256Mi" }
-            limits:   { cpu: "1",    memory: "1Gi" }
+            requests: { cpu: "100m", memory: "512Mi" }
+            # Headroom bump: the work-queue OOM is fixed in code, but a deep
+            # search still pulls up to 15000 Crossref records into RAM and the
+            # accumulating result JSONs are loaded whole. 2Gi gives margin so a
+            # big search isn't OOM-killed. Tune down if the node is tight.
+            limits:   { cpu: "1",    memory: "2Gi" }
      volumes:
        - name: doi
          nfs:
Author	SHA1	Message	Date
gitea	5d13843a99	conjurer: give librarian 60s termination grace for checkpointing The librarian now checkpoints an in-progress search on SIGTERM and exits within CONJURER_LIBRARIAN_GRACEFUL_TIMEOUT (default 45s). Raise k8s terminationGracePeriodSeconds to 60 so that graceful checkpoint isn't cut short by SIGKILL - otherwise the long DB scan restarts instead of resuming. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>	2026-08-02 22:11:05 +02:00
argocd-image-updater	7a33250d98	build: automatic update of conjurer updates image gitea/conjurer-librarian tag '8e18071b' to '40605b95'	2026-08-02 17:39:36 +00:00
gitea	d94ad4a0dc	conjurer: bump librarian memory limit 1Gi -> 2Gi (headroom) The work-queue OOM is fixed in code (bounded queue), but a deep search still loads up to 15000 Crossref records and the accumulating result JSONs into RAM. Give 2Gi of headroom so a big search isn't OOM-killed; raise the request to 512Mi to match its real baseline. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>	2026-08-02 19:36:14 +02:00
argocd-image-updater	791e745865	build: automatic update of conjurer updates image gitea/conjurer-librarian tag '44b7298a' to '8e18071b' updates image gitea/conjurer-bot tag '44b7298a' to '8e18071b'	2026-08-02 17:05:31 +00:00
argocd-image-updater	9fa72af1bf	build: automatic update of conjurer updates image gitea/conjurer-librarian tag '44b7298a' to '8ac68df1' updates image gitea/conjurer-bot tag '44b7298a' to '8ac68df1'	2026-08-02 17:03:28 +00:00
argocd-image-updater	abbba17dd0	build: automatic update of conjurer updates image gitea/conjurer-librarian tag '5d321f2f' to '44b7298a' updates image gitea/conjurer-bot tag '5d321f2f' to '44b7298a'	2026-08-02 15:33:09 +00:00