Tom Lane wrote:
I went through this thread again, and noticed something that no one
seems to have remarked on at the time: the vmstat numbers near the
bottom of this post
http://archives.postgresql.org/pgsql-general/2008-01/msg00161.php
show close to 100% I/O wait time (either that or 50% idle 50% I/O wait,
which I suspect is an artifact). We subsequently concluded that the
"SELECT" side of the INSERT/SELECT command is not where the problem is,
so all the cycles are going into the actual row insertion part.
I don't know of any reason to think that insertion is slower in 8.3
than it was in 8.2, and no one else has reported anything of the sort.
So I'm leaning to the idea that this suggests some kind of
misconfiguration of the disk setup in Clodoaldo's new server. There
was some earlier discussion about not having the RAID configured right:
http://archives.postgresql.org/pgsql-general/2008-01/msg00169.php
I'm thinking it's still not right :-( but have no real expertise
at diagnosing such problems. Anyone?
Per the thread.
# cat /proc/mdstat
Personalities : [raid1] [raid6] [raid5] [raid4] [raid0]
md2 : active raid1 sda1[0] sdb1[1]
224768 blocks [2/2] [UU]
md0 : active raid0 sda2[0] sdb2[1]
4497920 blocks 256k chunks
md1 : active raid1 sda3[0] sdb3[1]
310094592 blocks [2/2] [UU]
I don't see anything immediately wrong here. md0 is raid 0 and probably
his swap. md2 is probably boot and md1 is probably / . All volumes show
online.
If we are sure that this issue is apparent actual row insertion it
should be easy to duplicate. I will do some testing on my box which is
similar. I have 2gig of ram, dual core and a raid 1 with SATA drives.
Sincerely,
Joshua D. Drake
regards, tom lane
---------------------------(end of broadcast)---------------------------
TIP 5: don't forget to increase your free space map settings
---------------------------(end of broadcast)---------------------------
TIP 5: don't forget to increase your free space map settings
