huge strange io activity for no reason on raidz

nimmen

nimmen

hello, i have a problem with io activity
here is info:

7.1-STABLE FreeBSD 7.1-STABLE #5

root@[gw]:: zpool status
pool: pool
state: ONLINE
scrub: scrub completed with 0 errors on Thu Mar 19 16:52:19 2009
config:

NAME STATE READ WRITE CKSUM
pool ONLINE 0 0 0
raidz1 ONLINE 0 0 0
ad4 ONLINE 0 0 0
ad8 ONLINE 0 0 0
ad10 ONLINE 0 0 0
ad12 ONLINE 0 0 0

errors: No known data errors

root@[gw]:: zpool iostat
capacity operations bandwidth
pool used avail read write read write
---------- ----- ----- ----- ----- ----- -----
pool 1.31T 2.32T 33 14 3.88M 555K


top shows:
95.46% [irq20: atapci1]

CPU: 1.0% user, 0.0% nice, 6.2% system, 44.1% interrupt, 48.7% idle

root@[gw]:: iostat -xdw 1
extended device statistics
device r/s w/s kr/s kw/s wait svc_t %b
ad4 29.8 9.6 1311.3 187.1 0 23.3 4
ad8 29.9 9.1 1311.4 187.1 0 35.4 5
ad10 29.8 9.5 1319.7 186.9 0 24.3 4
ad12 29.7 10.4 1315.4 184.4 0 35.0 6
extended device statistics
device r/s w/s kr/s kw/s wait svc_t %b

ad4 0.0 0.0 0.0 0.0 0 0.0 0
ad8 0.0 0.0 0.0 0.0 0 0.0 0
ad10 0.0 0.0 0.0 0.0 0 0.0 0
ad12 0.0 0.0 0.0 0.0 0 0.0 0

gstat shows:

0 0 0 0 0.0 0 0 0.0 0.0| ad4
0 0 0 0 0.0 0 0 0.0 0.0| ad8
0 0 0 0 0.0 0 0 0.0 0.0| ad10
0 0 0 0 0.0 0 0 0.0 0.0| ad12

atapci1: <nVidia nForce MCP55 SATA300 controller> port 0x9f0-0x9f7,0xbf0-0xbf3,0x970-0x977,0xb70-0xb73,0xdc00-0xdc0f mem 0xfe02d000-0xfe02dfff irq 20 at device 5.0 on pci0
atapci1: [ITHREAD]
ata2: <ATA channel 0> on atapci1
ata3: <ATA channel 1> on atapci1

ATA channel 2:
Master: ad4 <Hitachi HDT721010SLA360/ST6OA31B> Serial ATA II
Slave: no device present
ATA channel 3:
Master: no device present
Slave: no device present

that is quite strange.
i have no idea why iostat first time shows huge activity and all next reports (using -w) shows zero activity, and actually it is zero activity, im 100% sure that there is nothing going on (no reads or writes), but zpool iostat shows activity too. gstat shows correct things, but top -S shows huge load on atapci1 and interrupts.

some loader.conf info:

vm.kmem_size="512M"
vm.kmem_size_max="1024M"
vfs.zfs.arc_max="256M"
kern.maxfiles="25600"
kern.maxfilesperproc="25600"

i had to change maxfiles because i ran out into limit few days ago. i have 2gb ram, it is home router, fileserver box.
CPU: AMD Athlon(tm) 64 X2 Dual Core Processor 4600+ (2411.13-MHz 686-class CPU)
 
i dont think that this will help (i think problem is in kernel or zfs or hardware(might be but im quite sure it is not)), output:

31 root -64 - 0K 8K WAIT 0 35.6H 90.38% [irq20: atapci1]

(pid =31)

root@[gw]:: pstree 31
--- 00031 root [irq20: atapci1]
root@[gw]:: pstree -p 31
root@[gw]::
 
This looks like a nForce4 chipset, those are pretty buggy, which may or may not have something to do with your problem.

Did you try rebooting the machine?
 
yes it is nforce4 chipset,
i will try rebooting some time later, but i thought maybe this problem would help improve nforce4 driver or some other issues related to this problem in freebsd :)
that problem does not affect me for now it would suffice even 1/20 of the CPU i have for my current needs, but 1 core is very busy other is idle, just strange io activity, i have this server for almost a year and saw this problem after doing zpool scrub.
 
You must log in or register to reply here.
Back
Top