5Stack

Appearance

System & Monitoring ​

Tools for watching the health and performance of the whole platform. (For configuring servers themselves; see Servers & Regions.)

System metrics ​

Live performance across your infrastructure:

  • Game server nodes: CPU and memory per node, with search, filters (only enabled / only online), sorting, and expandable history charts.
  • GPU nodes: per-node GPU and VRAM usage.
  • Services: the platform's internal services, with CPU/memory (warning and critical thresholds) and a jump to that service's logs.

System logs ​

Tail live logs for any service, api, web, game-server-node, demo-parser, hasura, typesense, timescaledb, redis, rustfs, mediamtx, with follow and timestamp toggles. Per-match server logs are available from the match's server tab (see Match server controls).

Database performance ​

A built-in database dashboard for keeping the backend healthy, covering:

  • Query performance: slow queries and execution times.
  • Connections and locks / transactions.
  • I/O & cache: buffer cache hit rates and disk I/O.
  • Index usage and storage sizes.
  • TimescaleDB hypertable stats.

Refresh on demand or auto-poll on an adjustable interval.

Backups ​

5Stack takes automatic database backups so your matches, stats, and configuration are protected. You can review backup status and history from the database dashboard.

Edge and storage checks ​

Two settings pages check the parts of the platform that live outside your cluster:

  • Settings → Application → Demos: Cloudflare delivery shows whether the Cloudflare Worker in front of your Backblaze bucket is answering and can read the bucket, and Storage check writes a test file to your bucket, reads it back (Write, Read), then fetches it through the worker (Edge) if you use one.
  • Settings → Application → Streaming: Edge relay (Cloudflare), shown when Playcast is on, says whether the relay worker is serving Playcast viewers from Cloudflare's cache.

Each reads Online, Needs attention, Offline or Not set up, with a Setup guide link. See Backblaze and Playcast Edge Relay for setting the workers up.

Server and node alerts ​

Nodes going offline, regions going down, and dedicated servers that stop answering are reported to administrators in the panel and on the Discord support webhook. See Status notifications for exactly when.

Game servers that crash leave a crash report on their node. See Game Server Crashes.

Updates ​

When a newer image is published for one of the panel's services, administrators see System Update Available listing the services. A rebuild that changes nothing no longer shows up there as a pending update.

Render queue ​

The render queue (administrator only) shows every clip and highlight render job and its progress. Pair it with GPU Nodes to see which node each job is running on. If renders back up. Check GPU availability and the "pause renders during live matches" setting on the Highlights page.