VAMI's pre-update check often refuses a patch or upgrade with a warning that /storage/archive is low on space. The partition holds archived PostgreSQL WAL history, and VCSA's archiver fills it close to the cap on purpose — so the warning is normal and genuinely blocking. Prune older archives, re-run the precheck, ship it. Applies to VCSA 6.7 / 7.0 / 8.0.
Before you touch anything
- Take a file-based VCSA backup via VAMI — supported recovery path if anything goes sideways.
- Optional but cheap: snapshot the VCSA VM. Drop the snapshot once the upgrade is stable; never keep one across a later disk expansion.
Prune the archive
SSH into the appliance as root and confirm the culprit:
du -sh /storage/archive/* 2>/dev/null
df -h /storage/archive
You'll almost always see /storage/archive/vpostgres consuming the bulk of the partition. Delete archives older than a retention window — 40 days is a safe default, narrow it only if you need more headroom:
# Preview first
find /storage/archive/vpostgres -type f -mtime +40 | head
# Delete
find /storage/archive/vpostgres -type f -mtime +40 -delete
Confirm headroom came back:
df -h /storage/archive
A healthy result looks roughly like:
Filesystem Size Used Avail Use% Mounted on
/dev/mapper/archive_vg-archive 49G 3.5G 43G 8% /storage/archive
Finish up
Re-run the VAMI Pre-Update Check — the /storage/archive warning should clear and the patch/upgrade can proceed. Once it's verified stable, drop the safety snapshot.
Worth knowing
- The partition refills as WAL accumulates. This is recurring maintenance, not a one-time fix — fold it into your pre-patch checklist.
- "Near full" on
/storage/archiveis by design; restoring headroom is what makes the precheck happy, not what fixes an underlying problem. - Make sure the retention window you choose still meets whatever your environment expects for point-in-time recovery before you delete.