admin/stonks-oracle
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

fix: use longhorn-rwx storage for cross-node pipeline scheduling

Browse Source 
- StorageClass longhorn-rwx with hard NFS mount (no softerr)
- Prevents Remote I/O errors from share-manager startup race
- RWX allows steps to spread across all 4 cluster nodes
- podAntiAffinity ensures different workflows use different nodes
This commit is contained in:
Celes Renata
2026-04-20 03:23:54 +00:00
parent a2f76cea96
commit 9850dc45b1
1 changed files with 2 additions and 4 deletions
Download Patch File Download Diff File Expand all files Collapse all files
.woodpecker.yml
+2 -4
View File
@@ -1,10 +1,8 @@
# Woodpecker CI pipeline for Stonks Oracle
# Translates .github/workflows/build.yml into Woodpecker native format
# Images push to registry.celestium.life/stonks-oracle/<service>:<sha> and :latest
#
# Scheduling: Pipeline pods are spread across cluster nodes via agent-level
# podAntiAffinity (WOODPECKER_BACKEND_K8S_POD_AFFINITY). Resource requests
# below help the scheduler make balanced placement decisions.
# Storage: longhorn-rwx (RWX) for cross-node step scheduling
# Scheduling: podAntiAffinity spreads pipeline pods across cluster nodes
steps:
lint-python: