[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[suse-security] two issues: sshd error messages and adaptec scsi card lack of stability



Hi,

Suse 9.1 installed on workstations. Central authentication using NIS from
sun server.

Error messages in /var/log/messages :

Sep  1 14:15:52 PC sshd[7347]: error: Could not get shadow information for
USER
Sep 1 14:15:52 PC sshd[7347]: Accepted password for USER from
::IPv^address port 1022 ssh2


Not a major problem, but two minor questions:
1) why am I being told that the user is not in the shadow password file,
when authentication is then being provided by NIS?
2) why is the client IP address being reported in IPv6 format?

both are annoyances, but I would like to turn off the first and change the
second to IPv4.

================

This is slightly OT, but given the recent spate of posts about production
quality of Suse 9.1 Pro I thought I would post. I've already reported this
as a problem to both Adaptec and Suse, but Adaptec SCSI RAID cards are
fairly common (?).

Hardware: Intel Xeon single CPU 4u high rack mount computer. 5 SCSI HDD, 2
IDE HDD. One Adaptec 2120S SCSI RAID controller. Configure Adaptec RAID
controller for RAID 5 with all SCSI HDDs (array builds...).
fdisk
mkfs -t ext3 /dev/sda1
mkdir /data1
mount -t ext3 /dev/sda1 /data1
all fine.
Do a simple write performance test onto new partition:
time dd if=/dev/zero of=/data1/8gb bs=1024k count=8192
This usually works and gives performance of about 38MB/second (not great,
but better than with the 2.4 kernel which was down at about 15 MB/second
which was awful).
Now create some files on /data1 by copying kernel source to it:
cp -r /usr/src/linux /data1
frozen xterm. Can login to computer, but any process that touches the RAID
array will hang and is completely unkillable.

Errors in log are like this:
Sep 9 09:52:39 tcmis kernel: aacraid: Host adapter reset request. SCSI
hang ?
Sep 9 09:52:39 tcmis kernel: aacraid: Host adapter appears dead
Sep 9 09:52:39 tcmis kernel: scsi: Device offlined - not ready after error
recovery: host 0 channel 0 id 0 lun 0
Sep 9 09:52:39 tcmis last message repeated 99 times
Sep 9 09:52:39 tcmis kernel: SCSI error : return code = 0x6000000
Sep 9 09:52:39 tcmis kernel: end_request: I/O error, dev sda, sector
476263772
Sep 9 09:52:39 tcmis kernel: Buffer I/O error on device sda3, logical
block 935884
Sep 9 09:52:39 tcmis kernel: lost page write due to I/O error on sda3
Sep 9 09:52:39 tcmis kernel: Buffer I/O error on device sda3, logical
block 935892
Sep 9 09:52:40 tcmis kernel: lost page write due to I/O error on sda3
(LOTS)


Now this could be a hardware problem, but the Adaptec card thinks it is
happy, and it used to work with Redhat 9 (the 2.4 kernel that I mentioned
earlier)......
This card is not supported with Suse 9.1 by Adaptec on their website:
http://www.adaptec.com/worldwide/support/drivers_by_product.jsp?sess=no&language=English+US&cat=%2FProduct%2FASR-2120S&prodkey=ASR-2120S

, but it has been out for a while now...

on this same webpage you can get the source of the latest source for the
aacraid module (1.1.5) which is different to the source used to the build
the 2.6.5-7.108-default kernel.... it refused to compile against the
running kernel.


I would like to know if other people have this problem, of if you have a
working RAID array with an Adaptec hardware controller.


regards,


Mike Rose

-- 
Check the headers for your unsubscription address
For additional commands, e-mail: suse-security-help@xxxxxxxx
Security-related bug reports go to security@xxxxxxx, not here