grpc storage writes failed with 'database or disk is full' while service stayed running

[butterson1]

Summary

opensnitch-web logged write failures from the gRPC path while the service remained up:

2026-03-23T14:10:01+00:00 bee opensnitch-web[1586715]: 2026/03/23 14:10:01 [grpc] Failed to store connection for @: database or disk is full
2026-03-23T14:10:01+00:00 bee opensnitch-web[1586715]: 2026/03/23 14:10:01 [grpc] Failed to store connection for @: database or disk is full
2026-03-23T14:10:01+00:00 bee opensnitch-web[1586715]: 2026/03/23 14:10:01 [grpc] Failed to upsert DNS mapping for @: database or disk is full

Context

• Service status at audit time: active (running)
• Uptime: since 2026-03-22 23:17:18 UTC
• Audit window: last 24 hours ending 2026-03-24 03:00 UTC

Notes

Host disk does not appear fully exhausted at the filesystem level during audit:

Filesystem                         Size  Used Avail Use% Mounted on
/dev/mapper/ubuntu--vg-ubuntu--lv   98G   87G  6.1G  94% /

So this may be one of:

• SQLite reporting database or disk is full due to DB/page/temp-file limits rather than the whole filesystem being out of space
• a temp directory or DB directory hitting a different capacity/quota/permission constraint
• DB corruption or another SQLite-level condition surfacing as SQLITE_FULL

Suggested next steps

• log the DB file path and underlying SQLite error code/context when this happens
• check whether temp files are being created in a location with tighter limits than /
• verify DB growth/retention behavior so connection and DNS tables do not hit an internal ceiling
• consider surfacing a clearer operator-facing warning when persistence fails but the service keeps running