I don't see the lack of I/O. I user adaptek serial ata raid of 10 level.
These are my stats:
iostat -k:
Linux amd64-k8-smp-aac 05/17/06
avg-cpu: %user %nice %sys %iowait %idle
6.82 0.00 7.18 43.17 42.83
Device: tps kB_read/s kB_wrtn/s kB_read kB_wrtn
sda 2.83 124.85 112.76 935291081 844728062
sdb 230.44 158.58 88.06 1187928059 659633296
or
vmstat 1
procs -----------memory---------- ---swap-- -----io---- --system-- ----cpu----
r b swpd free buff cache si so bi bo in cs us sy id wa
3 1 7008 84520 90840 607468 0 0 428 674 2 3 7 7 43 43
3 0 7008 74372 90872 607572 0 0 176 514 2065 3397 25 29 31 14
1 1 7008 91548 90936 607848 0 0 200 701 2342 4052 27 29 31 12
2 0 7008 90200 91012 608112 0 0 308 688 2309 3252 26 30 25 18
1 1 7008 89024 91052 608344 0 0 208 417 1764 3300 26 29 31 16
1 1 7008 81424 91116 610524 0 0 252 469 3222 5392 30 29 35 6
1 0 7008 86068 91232 611156 0 0 564 2771 2269 3039 19 24 24 34
0 0 7008 88092 91280 611584 0 0 248 1193 2391 3759 24 30 27 18
1 2 7008 79012 91348 611992 0 0 216 882 2021 3606 27 28 28 18
vmstat 1 -p sdb2
sdb2 reads read sectors writes requested writes
594831358 1077172625 1009896504 4292763184
594831408 1077173177 1009896630 4292765057
594831470 1077173857 1009896754 4292766245
594831530 1077174473 1009896851 4292767134
594831592 1077175145 1009896951 4292768542
594831652 1077175809 1009897013 4292769329
594831693 1077176225 1009897129 4292770799
594831750 1077176849 1009897234 4292772061
594831824 1077177585 1009897343 4292773535
594831869 1077178049 1009897423 4292774331
594831932 1077178681 1009897533 4292775643
594831976 1077179153 1009897715 4292778159
594832027 1077179713 1009897815 4292779332
594832096 1077180481 1009897925 4292780889
/var/spool/exim4 is on sdb2
Is using of more fast storage needed here?
On Tue, May 16, 2006 at 16:25 -0500, Phil Brutsche wrote:
> roos wrote:
> > Hi. I have noticed in logs :
> >
> > 2006-05-16 09:27:46 1Fft0D-0002em-Kr Failed to get write lock for
> > /var/spool/exim4/db/name_of_the_transport.lockfile: timed out
> > 2006-05-16 09:27:46 1FdqWa-0004ml-4i Failed to get write lock for
> > /var/spool/exim4/db/name_of_the_transport.lockfile: timed out
>
> I have found that it means that the disk(s) /var/spool/exim4/db resides
> on don't have the I/O capacity to meet Exim's needs... ie using IDE
> drives when you should be using SCSI.
>
> In my situation the messages went away when I put the hints databases on
> a tmpfs ramdisk - for me it wasn't critical that they be saved between
> reboots, and faster disks weren't an option.
>
> --
>
> Phil Brutsche
> phil@???
>
> --
> ## List details at http://www.exim.org/mailman/listinfo/exim-users
> ## Exim details at http://www.exim.org/
> ## Please use the Wiki with this list - http://www.exim.org/eximwiki/
