Repository navigation
What's Changed
- Cloud18 peering: the instance logs in to a peer with its realm token, not its password (#2327) by @svaroqui in #2328
- Restart mail obeys the mute (intervention, alert disable) by @svaroqui in #2335
- backup: mydumper write-stall watchdog (#2249) by @svaroqui in #2333
- GUI: system-config locks, durable config writes, Setup reorganised in tabs (#2268) by @svaroqui in #2334
- Monitoring settings: General and Performance Schema tabs; fix the HTTP & API crash (#2268) by @svaroqui in #2337
- ProxySQL: keep a user set at runtime across restarts; save cases for four secret keys by @caffeinated92 in #2332
- Peer poll: reconnect after a peer restart, and say why a peer is not connected by @svaroqui in #2340
- Alert URLs are plain settings again: read back the encrypted form v3.1.47 and v3.1.48 saved (#2342) by @caffeinated92 in #2343
- secret_store.json is written by the save callback only (#2345, part of #2041) by @caffeinated92 in #2346
- testLogicalReseed R8: wait for the archive push of the backup before the reseed from it (#2347) by @caffeinated92 in #2348
- Keep Cloud18 (OIDC) sessions across a restart (#2341) by @svaroqui in #2344
- license-builder: deliver an online instance's licence plugin via its -pull repo (#2309) by @svaroqui in #2349
- decrypt: print the configuration files to stdout, root only; read-only config load by @svaroqui in #2355
- Orchestrator secret copy (OpenSVC + K8s) via the StoreWriter queue, with one-deep _OLD rotation history (#2356) by @svaroqui in #2357
- App checks on an interval per check, registered backup check with one bucket listing per folder (#2362) by @svaroqui in #2363
- MySQL 8.4 behind ProxySQL 3: client logins, and the default admin login by orchestrator (#2360) by @svaroqui in #2364
- Global settings: switch cloud18-sso-login live (#2367) by @svaroqui in #2368
- Cluster roles and ACL: decisions on the review points of #2303 by @svaroqui in #2336
- Galera: the virtual master moves by itself while a majority is up; a lost node can rejoin (#2350) by @svaroqui in #2351
- Logical reseed and rejoin: armed task, rejoin backup gate, load stops at its first error (#2365) by @svaroqui in #2366
- Cloud18: the admin assigns or clears a cluster's external SysOps / DBOps directly (#2371) by @svaroqui in #2372
- Physical reseed: prepare on the backup's own undo directory, one receiver per physical backup by @caffeinated92 in #2380
- External SysOps / DBOps: write every change to the security log (#2371) by @svaroqui in #2373
- Cloud18 quote: a partner that is down or has no cluster left for sale no longer breaks the quote by @svaroqui in #2386
- Kubernetes: align database memory cap with OpenSVC and remove DBJobs memory limits (#2369) by @caffeinated92 in #2384
- Config git push keeps agents.json after a metadata refresh; monitor-only apps refuse the remaining orchestrator writes by @svaroqui in #2382
- Kubernetes: database and proxy PVC sizes follow the prov-db-disk-size unit contract by @caffeinated92 in #2388
- Bump go.opentelemetry.io/otel/trace from 1.45.0 to 1.47.0 by @dependabot[bot] in #2392
- Bump go.opentelemetry.io/otel/sdk from 1.45.0 to 1.47.0 by @dependabot[bot] in #2391
- API users: rotate a password to ban a compromised user; credential stores at two levels (server, cluster) by @svaroqui in #2376
- dbjobs: on-premise jobs as a non-root user use the jobs directory and find the server in /usr/sbin by @svaroqui in #2393
Full Changelog: v3.1.48...v3.1.49