This post is more than 5 years old
8 Posts
1
22082
Hosts connecting to ~management Storage Group
Hello,
I have three ESXi hosts which are connected to a Clariion CX3-20. They have two HBAs each, and they currently see storage just fine on the primary HBA (via both SPs). However, they are not seeing storage on their secondary HBAs. I have triple-checked zoning on the FC switch and it looks okay. The secondary HBAs do indeed see the Clariion and are registered with it, but they are seeing the wrong storage group. Instead of seeing VMWare Storage, they are seeing ~management. The VMware admin just sees “LUN0” on his end. This seems to make sense if the HBAs are seeing ~management instead of the VMWare group. I have attached screenshots of one of the ESXi hosts where I am seeing the problem as well as an ESX host that is okay. The array is running 03.26.020.5.025. I'm not sure if I need to do something different on the Clariion or if this may be on the VMware end.
Problem host:
Okay host:
Thanks
Matt
kenn2347
542 Posts
0
February 17th, 2011 08:00
Yep you can just click reconnect. that was going to be my next step. guess i should have put that in the first post:) I just like to verify that all paths are enabled first.
kenn2347
542 Posts
1
February 17th, 2011 07:00
Enter engineering mode and the go to the storage group properties. On the host tab there is an advanced button. hit that nad make sure that all the initiators are checked for each host. If some were not checked, after you check them do a rescan on the esx host and make sure both HBA's can now see the same amount of targets.
it also looks like from the pictures above that that bottom host is not online? it isnt logged in. Just noticed that
matt_heff
8 Posts
0
February 17th, 2011 08:00
Good catch on my "okay" host - it's in mtce mode.
Question: I just found this post in which someone references just clicking "Reconnect". Is that a viable option? Does clicking Reconnect cause a dicsonnect/reconnect on the HBAs that are already connected?
https://community.emc.com/message/410613
Thanks
Matt
dynamox
2 Intern
2 Intern
•
20.4K Posts
1
February 17th, 2011 08:00
you can use reconnect ..it will not affect working HBAs.
tkjoffs
159 Posts
1
February 17th, 2011 10:00
Side note: Maintenance Mode in ESX does not put the host in a Logged Out state. If it is doing that, you have issues with your ESX install or the HBAs you are using. In Maintenance Mode you should be able to perform maintenance on the Data Stores, and browse them, etc. Please double check that host is not offline and if it is, you need to dig deeper into the SAN/Host configurations on both nodes to figure out what is up.
ElvirK
23 Posts
0
February 17th, 2011 11:00
Was these HBAs used with another host and another array? Is ther array S/N# in managment SG and your SG same? I have this issue and in my case I have host/HBAs zoned and registered with another array.
christopher_ime
2K Posts
1
February 18th, 2011 04:00
Please note that when a registered host is in a storage group and later new paths are registered, they are not automatically associated with the host in the storage group as you experienced. To update it, there are three options:
a) Remove host from storage group and promptly add it back
b) Select "Reconnect" from "Connectivity Status" when navigating to the host within the storage group
only touch the new paths...
(an exercise for the reader to research how to access)
Finally, take appropriate measures on the host to update the paths to LUN(s).
zys1
1 Message
0
March 22nd, 2011 16:00
Regarding this hosts connecting to ~managment storage group issue, the reconnect solved the problem but where can I find offcial documentation about it? Thanks.
dynamox
2 Intern
2 Intern
•
20.4K Posts
0
March 23rd, 2011 13:00
here is a small note
emc114677
kelleg
4.5K Posts
0
March 24th, 2011 07:00
Please see the following Solution - emc126341
glen
driskollt1
131 Posts
0
March 24th, 2011 08:00
It probably happened like this..
1. You zoned Fabric A host-to-storage
2. ESX server registered the HBA connection in Navi
3. You zoned Fabric B host-to-storage
4. Added host to storage group
5. ESX Server registered the HBA connection in Navi
Basically, you work too fast. Be slower
Before you add the host to a storage group - Just make sure all connections for that host are logged in and registered.
If you ever replace an HBA, add an additional HBA, change what front-end ports an HBA is connected to - You'll need to reconnect the host to the storage group.
It's good that this happens. It lets you verify that you did the right thing before you give those HBAs access to the storage.