DEV Community

Michael
Michael

Posted on • Originally published at gbase8.cn

Managing GBase 8a Sync Recovery with gcadmin recover: VC Prioritization and Pause/Resume

#gbase #database #数据库

The built‑in gcrecover service in GBase 8a handles the recovery of data synchronization events. With the gcadmin recover command, DBAs can dynamically control recovery behavior — pausing or resuming tasks, and setting per‑VC priority for specific event types — without restarting any services. This article walks through the available subcommands with practical examples and log interpretation.

Subcommands

Run gcadmin --help and look for the recover section to confirm your version supports this feature. The available operations are:

  • pause: Stop scheduling new recovery tasks; in‑flight tasks continue to completion.
  • continue: Resume normal recovery scheduling.
  • highpriority vc_name [event_type]: Assign priority to a specific VC. The optional event_type can be ddlevent, dmlevent, or dmlstorage.
  • show status: Display whether the service is continue (normal) or pause.
  • show highpriority: Display the current priority rule; returns none if not set.

Examples

Pause and Resume

Both commands are idempotent — running them multiple times won't produce errors.

[gbase@vm151 config]$ gcadmin recover pause
gcadmin recover pause success
[gbase@vm151 config]$ gcadmin recover pause
gcadmin recover pause success
[gbase@vm151 config]$ gcadmin recover show status
pause
[gbase@vm151 config]$ gcadmin recover continue
gcadmin recover continue success
[gbase@vm151 config]$ gcadmin recover continue
gcadmin recover continue success
[gbase@vm151 config]$ gcadmin recover show status
continue
Enter fullscreen mode Exit fullscreen mode

Setting and Cancelling High Priority

You can target a specific VC and event type, and later cancel the priority rule.

[gbase@vm151 config]$ gcadmin recover highpriority vc_a ddlevent
gcadmin recover highpriority vc_a ddlevent success

[gbase@vm151 config]$ gcadmin recover show highpriority
gcadmin recover highpriority vc_a ddlevent

[gbase@vm151 config]$ gcadmin recover cancel highpriority
gcadmin recover cancel highpriority success
[gbase@vm151 config]$ gcadmin recover cancel highpriority
gcadmin recover cancel highpriority success
[gbase@vm151 config]$ gcadmin recover show highpriority
none
Enter fullscreen mode Exit fullscreen mode

Interpreting gc_recover.log

The log file clearly reflects state transitions and priority changes.

Pause and continue events:

2026-02-11 16:05:57.934 [INFO ] <RECOVER-INFO>: refresh gcrecover status is pause
2026-02-11 16:06:50.154 [INFO ] <RECOVER-INFO>: refresh gcrecover status is continue
Enter fullscreen mode Exit fullscreen mode

Priority rule applied, then cancelled:

2026-02-11 16:08:21.377 [INFO ] <RECOVER-INFO>: get gcrecover high pri rule is vc_a&ddlevent
2026-02-11 16:08:48.273 [INFO ] <RECOVER-INFO>: get gcrecover high pri rule is none
Enter fullscreen mode Exit fullscreen mode

Because the state is persisted in gcware, restarting gcrecover does not lose the current setting. The log will immediately reflect the last known status (pause or continue) right after the service comes back up.

When to Use These Controls

In a multi‑VC gbase database environment, gcadmin recover highpriority lets you ensure that a critical VC's metadata or data events are restored first. The pause/continue mechanism is a practical tool for reducing recovery‑related resource consumption during peak business hours — pause during the day, continue at night.

With these commands, GBASE gives you fine‑grained operational control over cluster synchronization recovery, helping your gbase database stay balanced and responsive even when large recovery workloads are in play.

Top comments (0)