network storage router e1200

[Hardydrewandthenancyboy]

We have a network storage router e1200 which is connected to our MSL5000. From the logs I can see a lot of unexplained reboots, about once a month. see below.
This is interfering with our backups as the devices go offline until the backup server is rebooted.
Log message says assertion failed. Anyone any ideas whats causing this?

Product Network Storage Router
Firmware Level 5.6.78
Bios Version BIOS 4.23
Booter Version BOOTER 4.11
CPU PLD Version 0x4

Event log:

1811 02/02/2006 10:50 27d17h02m22.83s CHK COND with Sense Key=0x4, ASC=0x8, ASCQ=0x0 from SCSI LUN 0/1/0
1812 02/02/2006 10:50 27d17h02m23.01s CHK COND with Sense Key=0x4, ASC=0x8, ASCQ=0x0 from SCSI LUN 0/0/0
1813 02/02/2006 10:50 27d17h02m23.19s CHK COND with Sense Key=0x4, ASC=0x8, ASCQ=0x0 from SCSI LUN 0/0/0
1814 02/03/2006 01:06 28d07h17m39.07s Assertion failed: 0, file ../ps_driver.c, line 2925
1815 02/03/2006 01:06 28d07h17m39.07s Assertion failed (continued): 2925
1816 02/03/2006 01:07 0d00h00m42.89s New device is added to location 0/0/0
1817 02/03/2006 01:07 0d00h00m42.89s New device is added to location 0/1/0
1818 02/03/2006 01:07 0d00h00m43.01s Unit restart and initialization, Firmware Version: 5.6 Build Level: 5.6.78
1819 02/03/2006 01:07 0d00h00m43.77s FC Port 0 Link is UP.

[Steven Clementi]

Kevin:


Is your SAN switch zoned in such a way that the NSR is in any one particualer zone which also has several hosts and/or storage arrays in it?


Steven

[Mark Poeschl_2]

Typically, your NSR should be in a zone that contains only itself and the host(s) that will be talking to the tape library. It should NOT contain any storage arrays.

[Steven Clementi]

And to take Marks' comment a bit further...


Single Initiator zoning should be applied, same as Storage Arrays. 1 zone for EACH server that needs to access the Tape Library totally separate from the 1 zone that server uses to access it's storage array.

Yes, it is a bit more management work, but at least you won't make the mistake of taking down 20 servers with one change to a zone. (potentially at least).


Steven

[Hardydrewandthenancyboy]

Our san switch is in it's default configuration. (HP StorageWorks SAN Switch 2/16 PN#AA978A)

There are a couple of servers, the nsr, an msa and an msl all fibre connected. I would guess they are all in one zone.
Can you point me at some documentation on zoning this switch?
Thanks a lot, -k

[Hardydrewandthenancyboy]

I've logged in an confirmed that the fabric switch is still at factory settings. Considering that the fabric is in use by several servers etc, is it possible to create these zones without incurring downtime?

[Uwe Zessin]

Well, if you do it correctly, yes.
The problem is that zoning is "all or nothing". By that I mean that you either run without zoning or need to set up everything at once and correctly before you activate the zoning configuration, else you can disrupt the communication.

While I have been working with zoning for quite some years now, I know that *I* would not risk to make an initial setup during rush hours unless really, truly, absolutely necessary.

[Steven Clementi]

The process is not so bad if your working with dual fabrics since you usually have availability on both fabrics, you can configure one at a time, wait to see if there is anything wrong I.E. Check paths to storage, access, etc. If all is ok, then configure the second fabric.


I have seen Storage Area Networks crawl to their knee's due to having no zoning configured. Odd things happen usually, but I have never seen anything "dangerous". Log files fill up with warnings and errors, loss of access to disk/tape luns occasionally with no explaination, etc.

If you post a capture of your switch info, we can possibly help you configure the initial zoning. Even if your going to wait until the weekend or at night or whenever... having some input would probably be great.

Telnet to the switch and capture the following commands:

switchshow, zoneshow, fabricshow, cfgshow, version

Or in the web gui on the switch, there is a "report" button on the admin page. Save to html and post here if you can/want.


Steven

[Hardydrewandthenancyboy]

Password:
swd77:admin>
swd77:admin>
swd77:admin> switchshow
switchName: swd77
switchType: 26.2
switchState: Online
switchMode: Native
switchRole: Principal
switchDomain: 1
switchId: fffc01
switchWwn: 10:00:00:05:1e:34:38:55
zoning: OFF
switchBeacon: OFF
Port Media Speed State
=========================
0 id N2 Online F-Port 10:00:00:00:c9:3b:85:fb
1 id N2 Online F-Port 50:08:05:f3:00:0f:92:01
2 id N2 Online F-Port 50:08:05:f3:00:0f:92:09
3 id N2 Online F-Port 10:00:00:e0:02:22:ca:4f
4 id N2 No_Light
5 id N2 Online F-Port 10:00:00:00:c9:3b:82:3c
6 id N2 Online F-Port 21:00:00:e0:8b:80:09:00
7 id N2 Online F-Port 21:00:00:e0:8b:1f:e1:f3
8 id N2 No_Light
9 -- N2 No_Module
10 -- N2 No_Module
11 -- N2 No_Module
12 -- N2 No_Module
13 -- N2 No_Module
14 -- N2 No_Module
15 -- N2 No_Module
swd77:admin> zoneshow
Defined configuration:
no configuration defined
Effective configuration:
no configuration in effect
swd77:admin> fabricshow
Switch ID Worldwide Name Enet IP Addr FC IP Addr Name
-------------------------------------------------------------------------
1: fffc01 10:00:00:05:1e:34:38:55 10.77.77.77 0.0.0.0 >"swd77"
swd77:admin> cfgshow
Defined configuration:
no configuration defined
Effective configuration:
no configuration in effect
swd77:admin> version
Kernel: 2.4.19
Fabric OS: v4.2.0b
Made on: Fri Apr 2 01:01:17 2004
Flash: Thu Jan 1 00:01:44 1970
BootProm: 4.1.0
swd77:admin>

[Hardydrewandthenancyboy]

Thanks, Config above. Note that we have only 1 fabric. What you say makes sense in that we have noticed sluggish transfer rates.
- k