MSA Storage
1748211 Members
4771 Online
108759 Solutions
New Discussion юеВ

Re: Event Source: RAIDISK

 
Dean_71
Occasional Contributor

Event Source: RAIDISK

I have an EVA 5000 SAN. Thursday morning one of my DL 580s, a node in my SAN attached MS 2003 Server Ent cluster, locked up. Cluster failover of this node resources did not occurr and SAN volumes owned by the node could not be accessed. Also, a demmand snapshot of one of the volumes owned by the node would not work. Although I could enter my logon information at the servers logon screen, I could not logon to the server.

I ran the Emulex HBAnywhere on another SAN attached server and the hung server's LP952 HBAs were online. All the port lights on the switches were green. A power cycle of the server resulted in gaining access to it.

I experienced the above symphtoms when a server in a two node test cluster with single HBAs had its HBA fail. However, this production server has dual HBAs.

I found the following envent:

Event Type: Error
Event Source: RAIDISK
Event Category: None
Event ID: 1026
Date: 3/9/2006
Time: 12:50:37 AM
User: N/A
Computer: xxxxxx
Description:
The Driver has detected a path failure to Subsystem ID 50001FE150008D60. Dump Data 0 contains the Phys Path Info < Port | Bus | Target | Lun >. Dump Data 1 contains the Driver Status. Dump Data 2 contains the HBA Slot Number (ffffffff if unavailable). Dump Data 3 contains extended Driver Status.
Data:
0000: 0f 00 10 00 02 00 60 00 ......`.
0008: 00 00 00 00 02 04 00 c4 .......├Д
0010: 00 00 00 00 00 00 00 00 ........
0018: 00 00 00 00 00 00 00 00 ........
0020: 00 00 00 00 00 00 00 00 ........
0028: 01 01 00 03 46 4f 44 00 ....FOD.
0030: 05 00 00 00 00 00 00 09 ........

I found the ID 50001FE150008D60 in the Emulex HBAnywhere, Target Mapping Tab, WWNN column for the servers HBAs. Also, all my other SAN attached servers have this ID.

Anyone have any ideas what caused this event error and did it result in my server locking up?

Thanks
6 REPLIES 6
Uwe Zessin
Honored Contributor

Re: Event Source: RAIDISK

50001FE150008D60 is the World Wide Node Name (WWNN) of the EVA5000, it is not an ID of any server.

According to the event log, the server had at least one path failure, but of course that should not cause a hang.

> Also, a demmand snapshot of one of the volumes owned by the node would not work.

Any error messages? Anything in the EVA event log?
.
Dean_71
Occasional Contributor

Re: Event Source: RAIDISK

On the SAN, I checked the Management Agent, Controller, and Controller Termination Event Logs. Everythign is check mark green.

I checked the event logs of my FC switches. On 9 Mar 06, switch one changed to a marginal/warning state for 5 sec starting at 01:27:22. Switch two changed to a marginal/warning state starting at 01:34:24 and ending at 01:35:54. I was unable to determine the current time on these switches to see if there is any time differential between the switches's and the server's event logs.

Anywhere else I can look?
Thomas Kreyche
New Member

Re: Event Source: RAIDISK

Did you get any further information about this? I just had a nearly identical event on my cluster EVA3000.

Thanks, Tom Kreyche
CA1428489
New Member

Re: Event Source: RAIDISK

Did you find any resolution to this?

We are having issues with our server locking up until a power cycle is initiated.

We are just using one server (not a cluster)

HP bl20p G3 attached to a EVA 5000

we get the same event in the event viewer.
Gwen Ketelslagers
New Member

Re: Event Source: RAIDISK

Hello,

We are experiencing exactly the same problems as described above...
Hardware:

EVA3000
BL20pG3 Blades booting from SAN
Windows 2003 Cluster

A HP techn. visited us today, and after long investigation he discovered a problem with Securepath. The diag. display returned -1 for both HBA's, and this should be 1 or 0.(according to the HP techn.) He also made some phone calls to various level 3 Helpdesk workers, and it appears that HP develop. knows more about this issue, but are refusing to give more information...

Strange thin about this is:

We always had these issues (raidist and ftdisk warnings upon booting a blade) but this never dit result in an error or anything else. Since a week we found out that our cluster is trying to fail over very often due to disk access problems. The non-clustered servers also have these eventlog errors, but they do not appear to have any real problems... yet..
Now why is this suddenly happening to a lot of people at the same time...
Looks to me that HP better start looking what is causing these problems.

Keona Smith
New Member

Re: Event Source: RAIDISK

We are experiencing the same problems on our SAN. Our SAN supports Windows 2003 and the Solaris 8 OS. We are using Secure Path 4.0 SP2. We have noticed that when a path fails it does not failover to our 2nd fabric that was built for redundancy. We are not sure if the RAIDISK errors is causing a latency issue between the servers and it's LUNs but we would most defintely like to know why the RAIDISK errors are occuring (excessively) within our event logs. Any help, answers, and guidance will be greatly appreciated.