Date: Thu, 8 Dec 2005 07:21:26 +0300 From: GreenX <freebsd@azimut-tour.ru> To: freebsd-stable@freebsd.org Subject: Primise and Intel RAID dont work Message-ID: <1693783100.20051208072126@azimut-tour.ru> In-Reply-To: <996076298.20051207170157@azimut-tour.ru> References: <64682545.20051207100535@azimut-tour.ru> <996076298.20051207170157@azimut-tour.ru>
next in thread | previous in thread | raw e-mail | index | archive | help
Hi PPL!
I have a computer with parent payment ASUS P4C800E-DELUXE.
OS FreeBSD 6.0-stable (CVS from 21.nov.2005) is established.
It is established five hard disks:
- One system on ICH5 ata100 Maxtor 6Y200P0
- And four on Promise RAID 0+1 WDC WD3200J (D/B)
All fine worked (except for em0, but also he has normally earned
after
updating)
while I have not wanted to establish two more hard disks (Maxtor
6Y080M0) on Intel in Stripe RAID.
Hard disks have been put - the system of them has found out
(the truth and known all bug "failed to enable memory mapping"),
and I with them was more did than nothing (did not format, did not
break - there was no time).
In three days of job the system has given out me the following:
ad4: req=0xc630f190 SETFEATURES SET TRANSFER MODE semaphore
timeout!!
DANGER Will Robinson!!
This warning has been given many times.
And ad4 it is established on Promise!
After reboot smartd has given out the warning
smartd [767]: Device:/dev/ad10, 1 Currently unreadable (pending)
sectors
smartd [767]: Device:/dev/ad10, 26 Offline uncorrectable sectors
(ad10 the first disk in Stripe RAID on Intel MATRIX)
After that has made CVSUP and has put for the night rebuild the
world.
In the morning has seen - to warrning
ad4: req=0xc90efd48 SETFEATURES SET TRANSFER MODE semaphore
timeout!!
DANGER Will Robinson!!
It was added:
kern.maxfiles limit exceeded by uid 100, please see tuning (7).
Has install the world, has spat and has disconnected these both
HDDs
from the controller
(itself Matrix has remained switched on)
Has passed three more days and yesterday has received the new
letter:)
kernel: DOH! ata_alloc_request failed!
kernel: FAILURE - out of memory in ata_raid_init_request
last message repeated 4 times
kernel: g_vfs_done ():ar0s1d [WRITE (offset=554081878016,
length=49152)] error = 5
kernel: g_vfs_done ():ar0s1d [WRITE (offset=554081730560,
length=12288)] error = 5
kernel: g_vfs_done ():ar0s1d [WRITE (offset=554082091008,
length=65536)] error = 5
kernel: g_vfs_done ():ar0s1d [WRITE (offset=554081927168,
length=16384)] error = 5
kernel: g_vfs_done ():ar0s1d [WRITE (offset=554081943552,
length=16384)] error = 5
syslogd: kernel boot file is/boot/kernel/kernel
kernel: panic: kmem_malloc (4096): kmem_map too small: 172470272
total
allocated
Then has switched off also the controller, has again updated system
and I wait some more days:)
Immemorial Russian question - "what to make?",
I think here it is inappropriate, and so all is clear
Remarks on system: it is switched on HHT (ACPI, SMP) and DEVICE
POLLING,
the network interface em0 and Intel Matrix, both stand on one irq.
---------------
Gy... That that is written above, I wrote yesterday and for a long
time to wait it was not necessary
In the evening the system has given out me the following:
kernel: ar0: WARNING - mirror protection lost. RAID0+1 array in
DEGRADED mode
kernel: subdisk4: detached
kernel: ad4: detached
kernel: unknown: TIMEOUT - WRITE_DMA retrying (1 retry left)
LBA=73576960
kernel: unknown: TIMEOUT - WRITE_DMA retrying (0 retries left)
LBA=73576960
After reboot the first disk on Promise appeared in a array 1, and
other disks in array 2.
Has changed sheduler on old 4BSD and has removed from a kernel device
polling
References
Visible links
Hidden links:
1. mailto:freebsd@azimut-tour.ru
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?1693783100.20051208072126>