| Name | Modified | Size | Downloads / Week |
|---|---|---|---|
| Parent folder | |||
| coroot-arm64 | 2026-09-07 | 71.7 MB | |
| coroot-amd64 | 2026-09-07 | 75.6 MB | |
| 1.26.0 source code.tar.gz | 2026-09-07 | 51.3 MB | |
| 1.26.0 source code.zip | 2026-09-07 | 51.6 MB | |
| README.md | 2026-09-07 | 2.9 kB | |
| Totals: 5 Items | 250.3 MB | 0 | |
MySQL: deep, DBA-grade monitoring
Coroot now reads MySQL's own diagnostics alongside eBPF wire capture, so a latency spike comes with the reason attached.
New health checks
- Latency — ranks the likely cause: heaviest query, row-lock waits, long transactions, buffer-pool pressure, redo-log stalls, purge lag, table locks, or CPU-bound
- Connection errors — clients failing the handshake (bad credentials, TLS, timeouts), as distinct from hitting
max_connections - Galera replication — flow control pausing writes cluster-wide, certification conflicts, nodes without quorum or still joining
- Group Replication — members not
ONLINE, growing certification and applier queues, lost redundancy - Backups — overdue, failed, or stale Percona XtraDB Cluster backups, with schedule and PITR state
InnoDB internals
Buffer pool in bytes (used vs free) and hit rate, dirty pages, disk reads, row operations, transactions, row-lock waits and average wait time, disk I/O and fsyncs, redo log throughput and waits, deadlocks, lock-wait timeouts, and undo history list length.
Disk growth
Binary logs and InnoDB undo tablespaces are now tracked and reported as growth sources in disk investigations. Both grow independently of table data and routinely fill a data volume — something table-size monitoring alone misses.
Query performance
Long-running queries are attributed while they're still running, not only once they finish. Rows examined vs returned per query surfaces missing indexes.
Galera and Group Replication are detected automatically. Works across MySQL, Percona Server, Percona XtraDB Cluster, and MariaDB.
What's Changed
- add metricsTTL to the installation docs by @freshpr in https://github.com/coroot/coroot/pull/976
- Remove a redundant config from the docs by @freshpr in https://github.com/coroot/coroot/pull/977
- advanced mysql inspections by @def in https://github.com/coroot/coroot/pull/979
- tracing: scope excluded peer IPs to the cluster being queried by @def in https://github.com/coroot/coroot/pull/980
- fix dependabot alerts: bump go modules, front and docs dependencies by @def in https://github.com/coroot/coroot/pull/981
- Harden CI/CD supply chain: pin actions, least-privilege permissions, Dependabot by @andife in https://github.com/coroot/coroot/pull/974
- dependabot: disable PR creation by @def in https://github.com/coroot/coroot/pull/988
- rca: cap the incident time window at 4h by @def in https://github.com/coroot/coroot/pull/991
- logs: allow filtering by trace id by @def in https://github.com/coroot/coroot/pull/992
- logs: allow fetching up to 10k messages by @def in https://github.com/coroot/coroot/pull/993
New Contributors
- @andife made their first contribution in https://github.com/coroot/coroot/pull/974
Full Changelog: https://github.com/coroot/coroot/compare/v1.25.0...v1.26.0
