Hi list,
I'm having trouble with slow requests, they have a noticeable impact
on the performance. I'd like to find out, what the root cause is, I
guess there are a lot of possible causes. But I'll just describe what
I'm seeing and hopefully someone can give advices.
I just counted the occurences of "slow request" in the ceph.log in the
last couple of days, all OSDs are affected, but host3 stands out:
osd.0 host1 447
osd.1 host1 59
osd.2 host1 1570
osd.13 host1 164
osd.19 host1 25
osd.3 host2 618
osd.4 host2 190
osd.5 host2 240
osd.9 host2 70
osd.10 host2 129
osd.6 host3 4323
osd.7 host3 2116
osd.8 host3 2321
osd.12 host3 5238
osd.11 host3 2019
osd.14 host4 2869
osd.15 host4 2024
osd.16 host4 2399
osd.17 host4 869
osd.18 host4 660
Hosts 1 to 3 have also a monitor running and many other services.
Currently, host4 has nothing else to do than running its OSDs, and
although the devices are bcached (the others aren't yet), it also
shows a significant amount of slow requests. But our monitoring shows
that this server is almost bored.
But on host3 we experience load peaks of 18 and higher. This server
has not the best hardware and it has many other tasks, so we are
wondering if this server can cause all the slow requests. Is it
possible that if host3 has a high load or a higher transfer, it
produces slow requests which are somehow inherited by the other nodes?
How do you deal with these slow requests?
Thanks for any help!
Regards,
Eugen
--
Eugen Block voice : +49-40-559 51 75
NDE Netzdesign und -entwicklung AG fax : +49-40-559 51 77
Postfach 61 03 15
D-22423 Hamburg e-mail : eblock@xxxxxx
Vorsitzende des Aufsichtsrates: Angelika Mozdzen
Sitz und Registergericht: Hamburg, HRB 90934
Vorstand: Jens-U. Mozdzen
USt-IdNr. DE 814 013 983
_______________________________________________
ceph-users mailing list
ceph-users@xxxxxxxxxxxxxx
http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
