hashar added a comment.

  I have looked at it live by running strace profiling against `dockerd` while 
a container was being reaped:
  
  name=strace -C -e trace=file -f -p `pidof dockerd`
    % time     seconds  usecs/call     calls    errors syscall
    ------ ----------- ----------- --------- --------- ----------------
     94.50    0.221674           2    145394     11771 unlinkat
      3.41    0.008000        2667         3         1 umount2
      1.71    0.004000        1000         4           fchmodat
      0.29    0.000674           0     11782           openat
      0.09    0.000220           0     11774         3 newfstatat
      0.00    0.000000           0         4           renameat
    ------ ----------- ----------- --------- --------- ----------------
    100.00    0.234568                168961     11775 total
  
  And is not even a full deletion.  But essentially yeah the slowness is due to 
the file deletion taking ages. Some `unlinkat` calls even take 100 - 200 ms. We 
have too many files to delete and there is a lot of I/O contention.

TASK DETAIL
  https://phabricator.wikimedia.org/T265615

EMAIL PREFERENCES
  https://phabricator.wikimedia.org/settings/panel/emailpreferences/

To: hashar
Cc: Daimona, thcipriani, hashar, awight, kostajh, Lucas_Werkmeister_WMDE, 
Alter-paule, Beast1978, Un1tY, Akuckartz, Hook696, Kent7301, joker88john, 
CucyNoiD, Nandana, NebulousIris, Gaboe420, Giuliamocci, Cpaulf30, Lahi, Gq86, 
Af420, Bsandipan, GoranSMilovanovic, QZanden, LawExplorer, Lewizho99, 
Maathavan, _jensen, rosalieper, Liudvikas, Scott_WUaS, Wikidata-bugs, aude, 
Jdforrester-WMF, Mbch331, Jay8g
_______________________________________________
Wikidata-bugs mailing list
Wikidata-bugs@lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/wikidata-bugs

Reply via email to