fix: bump pal-e-docs memory limit 128Mi → 256Mi #183

Merged

forgejo_admin merged 1 commit from 182-fix-bump-pal-e-docs-memory-limit-to-prev into main

2026-03-15 19:09:00 +00:00

forgejo_admin commented

2026-03-15 16:05:59 +00:00

Owner

Summary

Bump pal-e-docs container memory requests from 32Mi to 64Mi and limits from 128Mi to 256Mi. Container was getting OOMKilled (2 restarts in 12h) under normal load with FastAPI + SQLAlchemy + pgvector embeddings.

Changes

k8s/deployment.yaml: requests.memory 32Mi → 64Mi, limits.memory 128Mi → 256Mi. CPU values unchanged.

Test Plan

YAML validated with python3 -c "import yaml; yaml.safe_load(open('k8s/deployment.yaml'))"
After merge, confirm ArgoCD syncs the new limits and pod restarts without OOMKill
Monitor kubectl top pod in pal-e-docs namespace to confirm memory stays within new limits

Review Checklist

Passed automated review-fix loop
No secrets committed
No unnecessary file changes
Commit messages are descriptive

Closes #182
plan-pal-e-platform — operational stability

## Summary Bump pal-e-docs container memory requests from 32Mi to 64Mi and limits from 128Mi to 256Mi. Container was getting OOMKilled (2 restarts in 12h) under normal load with FastAPI + SQLAlchemy + pgvector embeddings. ## Changes - `k8s/deployment.yaml`: `requests.memory` 32Mi → 64Mi, `limits.memory` 128Mi → 256Mi. CPU values unchanged. ## Test Plan - [x] YAML validated with `python3 -c "import yaml; yaml.safe_load(open('k8s/deployment.yaml'))"` - [ ] After merge, confirm ArgoCD syncs the new limits and pod restarts without OOMKill - [ ] Monitor `kubectl top pod` in pal-e-docs namespace to confirm memory stays within new limits ## Review Checklist - [x] Passed automated review-fix loop - [x] No secrets committed - [x] No unnecessary file changes - [x] Commit messages are descriptive ## Related - Closes #182 - `plan-pal-e-platform` — operational stability

forgejo_admin added 1 commit

2026-03-15 16:05:59 +00:00

fix: bump pal-e-docs memory limit 128Mi → 256Mi

ci/woodpecker/pr/woodpecker Pipeline was successful

Details

6ab0ccf4fc

Container was getting OOMKilled with 128Mi limit (2 restarts in 12h).
FastAPI + SQLAlchemy + pgvector embeddings need more headroom.

Closes #182

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

forgejo_admin commented

2026-03-15 16:06:14 +00:00

Author

Owner

Review: APPROVED

Diff reviewed: 1 file, +2/-2 lines.

Findings: None.

Only k8s/deployment.yaml modified
requests.memory: 32Mi → 64Mi (correct)
limits.memory: 128Mi → 256Mi (correct)
CPU values unchanged
No secrets, no extraneous changes
YAML validated
Commit message is descriptive with Closes #182

Ship it.

## Review: APPROVED **Diff reviewed:** 1 file, +2/-2 lines. **Findings:** None. - Only `k8s/deployment.yaml` modified - `requests.memory`: 32Mi → 64Mi (correct) - `limits.memory`: 128Mi → 256Mi (correct) - CPU values unchanged - No secrets, no extraneous changes - YAML validated - Commit message is descriptive with `Closes #182` Ship it.