Pg files not deep-scrubbed
Pages: 1 2
admin
2,930 Posts
October 5, 2020, 1:54 pmQuote from admin on October 5, 2020, 1:54 pmAs of version 2.5 we support storing the configuration centrally (monitor database) which is the newer way in Ceph, the ceph.conf file is imported in the database and becomes minimal to specify the ips of the monitors. The import occurs as part of the upgrade process.
As of version 2.5 we support storing the configuration centrally (monitor database) which is the newer way in Ceph, the ceph.conf file is imported in the database and becomes minimal to specify the ips of the monitors. The import occurs as part of the upgrade process.
Last edited on October 5, 2020, 1:55 pm by admin · #11
khopkins
96 Posts
October 7, 2020, 1:25 pmQuote from khopkins on October 7, 2020, 1:25 pmWell updated to 2.6.2 and decided to do a "ceph repair all", now I have 37 of these errors overnight, how do you cancel the operation? the settings for the sleep and thress hold has been put back to normal. For now the deep-scrub is disabled in the maintenance field but the graphs say otherwise.
Well updated to 2.6.2 and decided to do a "ceph repair all", now I have 37 of these errors overnight, how do you cancel the operation? the settings for the sleep and thress hold has been put back to normal. For now the deep-scrub is disabled in the maintenance field but the graphs say otherwise.
admin
2,930 Posts
October 7, 2020, 6:21 pmQuote from admin on October 7, 2020, 6:21 pmthe errors you get are these not scrub in time warnings or some true error ?
the errors you get are these not scrub in time warnings or some true error ?
Last edited on October 7, 2020, 6:22 pm by admin · #13
khopkins
96 Posts
October 8, 2020, 1:30 pmQuote from khopkins on October 8, 2020, 1:30 pmI was getting a few every day but not this much since the cluster was updated. They did go down some since last night after the change back to original settings. The cluster is running nothing but SAS 4tb drives, no journal or SSD. IOPS run about 400-500 on normal operations. Is this too slow?
I was getting a few every day but not this much since the cluster was updated. They did go down some since last night after the change back to original settings. The cluster is running nothing but SAS 4tb drives, no journal or SSD. IOPS run about 400-500 on normal operations. Is this too slow?
admin
2,930 Posts
October 8, 2020, 5:50 pmQuote from admin on October 8, 2020, 5:50 pmyou need to adjust the scrub speed in steps so it its not too fast or too slow,this depends on your hardware and how many objects are stored. please see the earlier replies.
you need to adjust the scrub speed in steps so it its not too fast or too slow,this depends on your hardware and how many objects are stored. please see the earlier replies.
Pages: 1 2
Pg files not deep-scrubbed
admin
2,930 Posts
Quote from admin on October 5, 2020, 1:54 pmAs of version 2.5 we support storing the configuration centrally (monitor database) which is the newer way in Ceph, the ceph.conf file is imported in the database and becomes minimal to specify the ips of the monitors. The import occurs as part of the upgrade process.
As of version 2.5 we support storing the configuration centrally (monitor database) which is the newer way in Ceph, the ceph.conf file is imported in the database and becomes minimal to specify the ips of the monitors. The import occurs as part of the upgrade process.
khopkins
96 Posts
Quote from khopkins on October 7, 2020, 1:25 pmWell updated to 2.6.2 and decided to do a "ceph repair all", now I have 37 of these errors overnight, how do you cancel the operation? the settings for the sleep and thress hold has been put back to normal. For now the deep-scrub is disabled in the maintenance field but the graphs say otherwise.
Well updated to 2.6.2 and decided to do a "ceph repair all", now I have 37 of these errors overnight, how do you cancel the operation? the settings for the sleep and thress hold has been put back to normal. For now the deep-scrub is disabled in the maintenance field but the graphs say otherwise.
admin
2,930 Posts
Quote from admin on October 7, 2020, 6:21 pmthe errors you get are these not scrub in time warnings or some true error ?
the errors you get are these not scrub in time warnings or some true error ?
khopkins
96 Posts
Quote from khopkins on October 8, 2020, 1:30 pmI was getting a few every day but not this much since the cluster was updated. They did go down some since last night after the change back to original settings. The cluster is running nothing but SAS 4tb drives, no journal or SSD. IOPS run about 400-500 on normal operations. Is this too slow?
I was getting a few every day but not this much since the cluster was updated. They did go down some since last night after the change back to original settings. The cluster is running nothing but SAS 4tb drives, no journal or SSD. IOPS run about 400-500 on normal operations. Is this too slow?
admin
2,930 Posts
Quote from admin on October 8, 2020, 5:50 pmyou need to adjust the scrub speed in steps so it its not too fast or too slow,this depends on your hardware and how many objects are stored. please see the earlier replies.
you need to adjust the scrub speed in steps so it its not too fast or too slow,this depends on your hardware and how many objects are stored. please see the earlier replies.