How to speed up / slow down deep-scrubbing
This document (000019684) is provided subject to the disclaimer at the end of this document.
SUSE Enterprise Storage 6
Configure *osd max scrubs* and *osd scrub load threshold* to do more scrubbing work, but be careful as it will have a huge impact on performance.
Get current configuration by running the following command on an OSD node:
# ceph daemon osd.$osdID config show | egrep "osd_max_scrubs|osd_scrub_load"
# ceph daemon osd.22 config show | egrep "osd_max_scrubs|osd_scrub_load"
Assuming that impacting cluster clients is of no concern.
If system load average is above .5, then increase osd_scrub_load_threshold,
For system load average see: uptime, cat /proc/loadavg, top, sar -q 1 5, w
if load average is above consider increasing "osd scrub load threshold=", but may want to check randomly through out the day.
salt -I roles:storage cmd.shell "sar -q 1 5"
salt -I roles:storage cmd.shell "cat /proc/loadavg"
salt -I roles:storage cmd.shell "uptime"
Otherwise increase osd_max_scrubs:
ceph tell 'osd.*' injectargs --osd_max_scrubs=2 --osd_scrub_load_threshold=3
# ceph daemon $osdID config show | egrep "osd_max_scrubs|osd_scrub_load"
# ceph daemon $osdID config diff
Values will vary on each cluster.
Monitor for desired results with "ceph -s", etc....
If need drop back values back down closer to defaults if the commands negatively impact the cluster.
Cluster may have client load.
This Support Knowledgebase provides a valuable tool for SUSE customers and parties interested in our products and solutions to acquire information, ideas and learn from one another. Materials are provided for informational, personal or non-commercial use within your organization and are presented "AS IS" WITHOUT WARRANTY OF ANY KIND.
- Document ID:000019684
- Creation Date: 14-Aug-2020
- Modified Date:14-Aug-2020
- SUSE Enterprise Storage
For questions or concerns with the SUSE Knowledgebase please contact: tidfeedback[at]suse.com