Yes, pool is growing (~30k obj/sec) also client io is 10k op/s wr, 20k op/s rd May be you right, may be PG in migration state. I was reweighted this OSD. For 761,718 nearfull is also not triggered. 705 nvme 0.91199 1.00000 913 GiB 743 GiB 613 GiB 8 KiB 130 GiB 170 GiB 81.39 1.30 46 up osd.705 12 nvme 0.91199 0.73999 912 GiB 744 GiB 609 GiB 12 KiB 135 GiB 168 GiB 81.55 1.31 45 up osd.12 651 nvme 0.91199 1.00000 913 GiB 749 GiB 619 GiB 12 KiB 130 GiB 164 GiB 82.04 1.31 45 up osd.651 759 nvme 0.91199 1.00000 913 GiB 760 GiB 624 GiB 8 KiB 136 GiB 153 GiB 83.23 1.33 45 up osd.759 667 nvme 0.91199 0.96999 913 GiB 762 GiB 633 GiB 8 KiB 129 GiB 151 GiB 83.41 1.34 46 up osd.667 720 nvme 0.91199 1.00000 913 GiB 762 GiB 626 GiB 12 KiB 136 GiB 151 GiB 83.45 1.34 47 up osd.720 2 nvme 0.91199 1.00000 912 GiB 763 GiB 626 GiB 12 KiB 137 GiB 149 GiB 83.66 1.34 47 up osd.2 712 nvme 0.91199 0.92999 913 GiB 770 GiB 632 GiB 12 KiB 138 GiB 143 GiB 84.32 1.35 46 up osd.712 38 nvme 0.91199 0.95999 912 GiB 773 GiB 639 GiB 12 KiB 133 GiB 139 GiB 84.71 1.36 48 up osd.38 761 nvme 0.91199 1.00000 913 GiB 778 GiB 639 GiB 8 KiB 138 GiB 135 GiB 85.19 1.36 48 up osd.761 718 nvme 0.91199 0.90999 913 GiB 783 GiB 643 GiB 12 KiB 139 GiB 130 GiB 85.73 1.37 48 up osd.718 696 nvme 0.91199 0.90999 912 GiB 837 GiB 689 GiB 4 KiB 148 GiB 75 GiB 91.79 1.47 47 up osd.696 ID CLASS WEIGHT REWEIGHT SIZE RAW USE DATA OMAP META AVAIL %USE VAR PGS STATUS TYPE NAME k
On 21 Apr 2021, at 21:21, Dan van der Ster <dan@vanderster.com> wrote:
Are you currently doing IO on the relevant pool? Maybe nearfull isn't reported until some pgstats are reported.
Otherwise sorry I haven't seen this.