This post is more than 5 years old
2 Intern
•
388 Posts
0
7664
January 21st, 2013 19:00
IP Address is incorrect (ESXi 5 Hosts) in Unisphere Manager ?
Hi,
For unknown reason, the IP addresses of those ESXi 5 Hosts connected to CX4-240 Storage Array has been changed from x.x.15.x to x.x.13.x
In this way, we are not able to update information for those ESXi Hosts. Moreover, one of them has status = "Unmanaged".
We have restart the Management Server for the Storage Array but still gets the same result. On the other hand, all ESXi Hosts are working properly - All of them are connected to LUNs assigned and VMs are running too.
We would like to seek your advice is there any way to change the IP address ? We have attempted to connect Host for x.x.15.x but it failed. I guess it is due to the HBAs are already logged in.
Your advice is osught.
christopher_ime
2K Posts
0
January 28th, 2013 22:00
I'm not entirely certain what you need to do to shuffle things around with minimal impact to your environment, but what gets passed to Unisphere via automatic host registration is the IP assigned to vmk0. We rely on that being able to access the SP ports and vice versa. I'll assume as is best practice, that the vMotion VMkernel subnet (curiously assigned to vmk0 in your environment) you specified is a separate vlan from all else which would explain much of what you reported in the early posts. Could you follow-up with VMware support first?
TonyJK2
2 Intern
•
388 Posts
0
January 21st, 2013 20:00
After restarting the Management Server, the status of that particular ESXi Host becomes Normal. Virtual Machines running on it are shown up properly.
However, the IP addresses are still incorrect.
Thanks
dynamox
9 Legend
•
20.4K Posts
1
January 21st, 2013 20:00
Had the same issue, ask your ESX guy to check local configuration on the host.
kelleg
4.5K Posts
1
January 22nd, 2013 08:00
ESX since version 4.x includes a Host Agent (similar to the EMC Host Agent) that pushes to the LAN IP address (and the failover mode setting) to the array on boot. Make sure that you do not have both the EMC Host Agent (maybe a hold over from an older pre-4.x version that's been upgraded to 5) and the VMware Agent installed.
glen
TonyJK2
2 Intern
•
388 Posts
0
January 22nd, 2013 15:00
Dear Glen,
After changing the ESXi Host VMKernel Settings, it does work for a while but failed again.
Yes, you are right. We upgrade the ESXi Host from 4 (with Navisphere Agent) to 5. Is there any way to find out whether EMC Host Agent still exists and how to remove it ?
Your advice is sought.
Thanks
dynamox
9 Legend
•
20.4K Posts
0
January 22nd, 2013 19:00
box was upgraded to 5, there is no rpm in ESXi
AnkitMehta
1.4K Posts
0
January 22nd, 2013 19:00
In order to uninstall the Navisphere Agent you may try the below steps:
rpm -qa | grep navi
rpm -e naviagent-
where
is the old versionrpm -e naviinitool*
ESX_uninstall.sh
- Ankit
dynamox
9 Legend
•
20.4K Posts
0
January 22nd, 2013 19:00
Don’t let us down Ankit ☺ ..i’ll ask my ESX admin in the morning (although we upgraded from ESXi 4 not ESX 4)
AnkitMehta
1.4K Posts
0
January 22nd, 2013 19:00
I did get a response from vExperts. They are on it, they are verifying it with the community, with time permiting one of them will update. Since its late night in EST, we might have to wait till morning!
AnkitMehta
1.4K Posts
0
January 22nd, 2013 19:00
dynamox , TonyJK : Okay I tried researching more but no luck with this and it has made me curious too so I have reached out to vExpert & vSpecialist Community. With time permitting, expect a response soon!
[I hope so]
AnkitMehta
1.4K Posts
0
January 22nd, 2013 19:00
Oh! Sorry my bad I forgot to check version applicable. The above is applicable till 3.5.x
Thanks for pointing it out!
kelleg
4.5K Posts
0
January 23rd, 2013 06:00
Check the Storage Group that contains the LUNs - make sure that one of the LUNs is marked as HLU 0 (Host LUN). Without a LUN 0 (that means you will have a LUNZ present) the hosts may try to register incorrectly. This does not specifically address the IP addresss change, but might be part of the problem.
You also may need to shutdown the hosts, remove the hosts from the Storage Group, de-register the hosts, then power-on the host and see if the IP address is correct. The try the other hosts (assuming there are more than one host).
glen
TonyJK2
2 Intern
•
388 Posts
0
January 23rd, 2013 17:00
Dear Glen,
Thanks for your advice.
Does the Host ID = LUN ID ?
For instance, in Unisphere, we get
ProdLUN [2, RAID5, ESX1 .... ]
and 2 is the Host ID ?
Thanks
dynamox
9 Legend
•
20.4K Posts
0
January 23rd, 2013 17:00
LUN ID is not Host ID, if you look at storage group properties, column on the right is Host ID. You can only change Host id when you add LUNs to storage group.
AnkitMehta
1.4K Posts
0
January 23rd, 2013 19:00
Adding to dynamox already mentioned about HostID.
Host ID is nothing but your HLU value i.e A host alias for a logical unit number (LUN) that exists on a CLARiiON storage array. When a LUN is added to a storage group for masking purposes, it is given a default HLU value, different from the LUN value, that is visible to the host. When a LUN is added to a storage group, it is given a default HLU value that is visible to the host. You can change the default value by selecting from a list of available values in the Available HLU column of this table/ though the CLARiiON Storage Group Configuration wizard. The change occurs after the storage group configuration task is executed.
As far as my understanding goes LUN ID refers to A unique hexadecimal number that identifies a group of one or more disks in a storage array, in one area of disk storage space, that looks like a single disk to the operating system. Synonymous with logical storage device, storage device, and LDEV.
I tried reaching out to VMWare vExperts and vSpecialists but I have not received any response as yet. I reckon, I will try read more about ESX4.x to ESX5 upgrade and issues related to it and other related information this weekend and if I found anything I will update this space.