- Community Home
- >
- Storage
- >
- Entry Storage Systems
- >
- MSA Storage
- >
- Re: MSA1000 failvover and HBA questions
Categories
Company
Local Language
Forums
Discussions
Forums
- Data Protection and Retention
- Entry Storage Systems
- Legacy
- Midrange and Enterprise Storage
- Storage Networking
- HPE Nimble Storage
Discussions
Discussions
Discussions
Forums
Forums
Discussions
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
- BladeSystem Infrastructure and Application Solutions
- Appliance Servers
- Alpha Servers
- BackOffice Products
- Internet Products
- HPE 9000 and HPE e3000 Servers
- Networking
- Netservers
- Secure OS Software for Linux
- Server Management (Insight Manager 7)
- Windows Server 2003
- Operating System - Tru64 Unix
- ProLiant Deployment and Provisioning
- Linux-Based Community / Regional
- Microsoft System Center Integration
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Discussion Boards
Community
Resources
Forums
Blogs
- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
тАО08-16-2006 03:24 AM
тАО08-16-2006 03:24 AM
MSA1000 failvover and HBA questions
I recently installed an MSA1000 (two controllers), two FC switches, and a few hosts (each with 2 Qlogic HBAs installed).
1) The goal is to have redundant data paths - I am not sure if my customer has purchased Secure Path. Will HP MPIO suffice?
2) Looks like everything is working, but for some reason, at some point SANSurfer Express has stopped seeing anything.
Basic setup is as follows:
ServerA:
HBA_1 connects to FCSwitch1
HBA_2 connects to FCSwitch2
ServerB:
HBA_1 connects to FCSwitch1
HBA_2 connects to FCSwitch2
MSA1000:
Controller1 connects to FCSwitch1
Controller2 connects to FCSwitch2
Zoning on the switches:
on FCSwitch1:
Created a Zone Set named for the cluster I will be installing.
Created 2 Zones
Zone1 members = ServerA_HBA1 and MSA1000_Connector1 are members (aliases created and used)
Zone2 members = ServerB_HBA1 and MSA1000_Connector1 are members (aliases created and used)
Activated this Zone Set.
on FCSwitch2:
Created a Zone Set named for the cluster I will be installing.
Created 2 Zones
Created 2 Zones
Zone1 members = ServerA_HBA2 and MSA1000_Connector2 are members (aliases created and used)
Zone2 members = ServerB_HBA2 and MSA1000_Connector2 are members (aliases created and used)
Activated this Zone Set.
So far, everything is looking good. I launch ACU, everything looks good. I created a few RAID arrays, Logical disks, assigned hot spares. All works... I used SSP, and for now, only allow ServerA to see the logical disks (until the cluster is installed and I am ready to install node2).
If I an using SANSurfer (from either host), I get an error taht says SANSurfer Express did not detect any storage subsystem.
I checked and dbl checked that there was only one instance of SANSurfer running.
I know I don't absolutely need it, but it can be a useful tool when diagnosing problems. I saw one other post that indicated a similar issue, but did not notice any resolve.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
тАО08-16-2006 11:06 PM
тАО08-16-2006 11:06 PM
Re: MSA1000 failvover and HBA questions
If you install MPIO you will need to remove SANsurfer software.
This version of MPIO is somewhat tricky to install, but i use it on my clusters. After the cluster is built, move all the clus resources to one node. On the other node open the ACU | SSP and uncheck one hba's access to the disks.
Install HP MPIO and reboot.
Opent the MPIO CLI fm start menu and type: hpdsm devices This will show the luns and a single path.
Add back the other HBA access in SSP and reboot the host again! (ugh)
Run the MPIO CLI again and you should see 2 paths.
Do the same steps on the other node after moving the clus resources.
The setup looks correct, However if i were zoning these clusters I would use the 'single initiator'zoning rule, so i would have 4 zones, each with a host_hba and an MSA port. Its easier to troubleshoot in the long run and minimizes the chances broadcasting SCSI resets from one host to another.
We tried using SANSurfer a couple years back for multipathing but did not fully test it and noone here had any insterest in do testing because we were using SecurePath, which is a great product. Actually, HP MPIO is based on SecurePath, if you look at the docs and the file listing you will see they use some Sp dlls.
goodluck
Are your HBAs 2214's? last year we had problems with them in clusters. This issue was on a reboot the host would lose the luns completely. There was no persistence. We went back to using the Emulex based HBAs that have never had any problems.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
тАО08-16-2006 11:26 PM
тАО08-16-2006 11:26 PM
Re: MSA1000 failvover and HBA questions
I think that is what I did... two zones per fibre switch for a total of four... each zone only has one HBA and one MSA controller.
The HBAs are QLogic 2940 I believe. PCI-X-133 cards.
Anyhow, after I posted yesterday, I started playing with MPIO and it seems to be installed and working. SANSurfer is also installed, just not working.
I have assigned 4 LUNs to my host, and filled the check boxes in SSP for each HBA on this node. In my Device Manager under Disk Drives, in addition to the locally attached hard drive, it lists 8 logical drives and 4 multi-path hard drives. I am guessing that is normal... Under Disk Management it only lists the 4 LUNs that were assigned to this host.
Ultimately, I was able to install the cluster without any issues, and both nodes are functioning just fine, so I guess my next set of tests will be to pull a fibre cable and see how the multi-path performs.
What I plan on doing to test is use the MPIO CLI and look at the path status, pull one FC cable, watch the cluster service (it should NOT failover yet), and then look at the MPIO CLI again for the path status.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
тАО08-16-2006 11:44 PM
тАО08-16-2006 11:44 PM
Re: MSA1000 failvover and HBA questions
The server OS is Win2K3 R2.
http://search.hp.com/query.html?cc=usтМй=en&qt=MPIO&la=en
Is the HP MPIO to be used one of the ones in the URL above?
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
тАО08-17-2006 02:13 AM
тАО08-17-2006 02:13 AM
Re: MSA1000 failvover and HBA questions
it should say something like that in Device Manager. It seems MPIO is correctly installed. For each presented LUN, you see two disk devices and one multipath device, which is as it should be. And in Disk Management you're OK, as well. If MPIO was not working, each disk would probably show up twice in Disk Management.
And the MPIO Framework is from Microsoft, which probably explains the driver version. Only the DSM (Device Specific Module) comes from HP.
Here's the link to the latest verion of the Basic MPIO DSM. The Full-Featured one only works with the MSA1500 with Active/Active firmware (not available for the MSA1000 yet).
http://h20000.www2.hp.com/bizsupport/TechSupport/SoftwareDescription.jsp?lang=en&cc=us&prodTypeId=18964&prodSeriesId=421492&swItem=co-40735-1&prodNameId=421495&swEnvOID=1005&swLang=13&taskId=135&mode=4&idx=0
Regards,
Stephen