LIP reset / LOOP down

[Philip Dunn_1]

Hi to all,
Wonder if anyone with some SAN experience can help?
We have some boot-from-SAN bl35p blades (running redhat 4u4) in an enclosure. Recently several of the blades have started behaving quite strange, giving read-only errors when writing to filesystems, LIP reset and LOOP down errors in syslog (see below) and in worst cases loosing published disks completely.
Has anyone ever come across anything like this before and could point me in the right direction of things to check in order to resolve this issue?

Many Thanks in advance.

Phil

# touch test
touch: cannot touch `test': Read-only file system

sample of /var/log/messages:
Mar 8 13:54:36 kernel: qla2300 0000:04:01.0: LIP occured (b5b5).
Mar 8 13:54:39 kernel: qla2300 0000:04:01.0: LIP reset occured (b5f5).
Mar 8 13:54:40 kernel: qla2300 0000:04:01.0: LOOP DOWN detected (2).
Mar 8 13:54:44 kernel: qla2300 0000:04:01.1: LIP reset occured (f700).
Mar 8 13:54:44 kernel: qla2300 0000:04:01.0: LIP occured (f8f7).
Mar 8 13:54:44 kernel: qla2300 0000:04:01.1: LIP occured (f700).
Mar 8 13:54:46 kernel: qla2300 0000:04:01.0: LOOP UP detected (2 Gbps).
Mar 8 13:54:46 kernel: qla2300 0000:04:01.1: LOOP UP detected (2 Gbps).
Mar 8 13:54:46 kernel: scsi(0) :Loop id 0x0082 is an EVA_XL (AA) device
Mar 8 13:54:46 last message repeated 2 times
Mar 8 13:54:46 kernel: scsi(0) :Loop id 0x0083 is an EVA_XL (AA) device
Mar 8 13:54:46 last message repeated 2 times
Mar 8 13:54:46 kernel: scsi(1) :Loop id 0x0081 is an EVA_XL (AA) device
Mar 8 13:54:46 last message repeated 2 times
Mar 8 13:54:46 kernel: scsi(1) :Loop id 0x0082 is an EVA_XL (AA) device
Mar 8 13:54:46 last message repeated 2 times

[Brian Mousseau]

I'm having the same problem. Please post an update if you have solved this.

[Philip Dunn_1]

Yes, Redhat have fixed this in an update 5.
http://kbase.redhat.com/faq/FAQ_85_9610

Hope it helps

Phil

[Patrick Terlisten]

Hello,

AFAIK is this a normal behaviour. I also seen this with Windows hosts, they report also errors to the event log. Two BL30/35 blades in a sleeve are bulding a FC-AL. This loop is connected to the fabric (fabric loop). This is needed because only 8 ports for outgoing fc-connects are available with a p-Class enclosure. But you can use it with up to 16 blades (BL30/35). While booting one blade in the sleeve, a LIP will be generated. This is noticed by the other blade. You can see this in you logs or eventlog.

Hope this helps.

Best regards,
Patrick