Originally posted by: stephen2615
I intend to place a call with HDS in the morning (as I don't want to sit at my desk all night waiting for a response) on this but perhaps someone else has seen this.
We have a Windows 2003 server that the users always want re-imaged due to some problem with the software they are trying to implement. So I have done this before without this issue.
Today the server was re-imaged and the server team gave it to me to put on HDLM (5.7) and other SAN related software. The server is connected to a USP with three LUN's through two fabrics.
After rebooting the server due to the drivers, etc, it refused to come up. It got partially up but thats about it. After much discussion we thought the image might be at fault so I decided to de zone the server from one fabric and when I did that, the server started as I normally wont re zone the fabric too close together. I had a look at the events and found lots of issues with HDLM and paths that were missing or invalid. I re zoned the server again and all the paths came on line except one LUN that I was able to manually bring online. All seemed ok. So I decided to reboot the server to make sure it was all ok again. Same thing happened again. I thought it might still be an issue with the image so we re-imaged it again and after installing HDLM, etc, it still had the problem.
So, I re-installed HDLM, no change. As this was the only server on the new USP ports, I thought it might be an issue with one port so I moved the server to other ports. No change.
So, I thought that perhaps the LUN's which are dual pathed might be some how corrupt so I pulled the LUSE LUNs out of the host group and remade them by mixing up the old 100 GB LDEV's. I formatted the new LUNs and rebooted the server again. No change.
One of the LU paths always seems to appear offline when it is presented. The other two LU's are ok. When I online the LU in HDLM, all seems ok. The LU that is offline initially is going through the fabric that I de-zone and this makes the server come up ok. Only one path has this issue. The other path is just fine.
When I check what HDLM is see by using:
dlnkmgr view -path or
dlnkmgr view -lu,
sometimes it is wrong and it wont refresh properly. It will have LU's that are not there anymore and are offline but dlnkmgr still shows them as being online.
So, I am totally confused as to what is causing this one LU to be a problem considering I released it and rebuilt it again as a LUSE from other LDEV's.
Does anyone have any ideas on what could be causing this? Please don't say upgrade to a later version of HDLM as I am still waiting for it and so far HDLM 5.7 has never caused us any problem. All the HBA's are running a recent and standard driver that hundreds of other servers are using. This server is exactly the same as the others as we have a standard SOE that works well with other servers.
This is the first time I have seen something like this and I have been doing this for years.