Rescan ESX

[ii00907]

Goodmorning,

i have a question for you.

Every day for processes of clone db, I use the scripts

rdmManagmement.pl for destroy old rdm ( LUN ) and Add new rdm ( LUN ) to VM

and

rescanAllHBAInCluster.pl for rescan HBA when i Remove or Add new LUN on ESX

the script works great, very thanks William Lam.

Now the question;

rescanAllHBAInCluster.pl , rescan all HBA for all host ( esx ) in a cluster and the time for rescan is "X".

I have a rule on DRS that the VM should reside always in a specific host, so as to make the rescan only for a single host whit command

vicfg-rescan --server xxxxxxx --username xxxxxx --password xxxxxxx vmhba1

vicfg-rescan --server xxxxxxx --username xxxxxx --password xxxxxxx vmhba2

and then, at the end of the process add Lun to the VM , perform the rescan all HBAs for hosts to Internal Cluster rescanAllHBAInCluster.pl,

with this process rescan times a single host is less than 4 Host.

but with the command vicfg-rescan does not immediately see the new LUNs on the host, and therefore, subsequently, add the new Lun with

rdmManagmement.pl goes wrong.

I have insert a sleep between vicfg-rescan and rdmManagmement.pl , but it is not always ok.

I was wondering if there is a script, performing as rescanAllHBAInCluster.plrescanAllHBAInCluster.pl, but for specific host "rescanAllHBAInHost.pl.

Sorry for my bad English and thanks for your help.

ii00907

Ivan

[lamw · Accepted solution]

Here is a quick hack to get rescanAllHBAInCluster.pl to support a single ESX(i) host rescan - http://vghetto.svn.sourceforge.net/viewvc/vghetto/scripts/rescanAllHBAInCluster.pl?view=log

You can specify either --cluster or --host:

./rescanAllHBAInCluster.pl --server reflex --cluster Primp-Prod-Cluster

Scanning cluster: Primp-Prod-Cluster ...
   Rescan all HBAs successful for host himalaya.primp-industries.com.
   Rescan for new VMFS volumes successful for host himalaya.primp-industries.com.
   Refresh storage information successful for host himalaya.primp-industries.com.
Scanning cluster complete!

./rescanAllHBAInCluster.pl --server reflex --host himalaya.primp-industries.com

Scanning host: himalaya.primp-industries.com ...
   Rescan all HBAs successful for host himalaya.primp-industries.com.
   Rescan for new VMFS volumes successful for host himalaya.primp-industries.com.
   Refresh storage information successful for host himalaya.primp-industries.com.
Scanning host complete!

I quickly tested this, but should give you what you're looking for.

[ii00907]

Hi William,

many many many thanks!!!!

Despite my bad English, you have understood what i needed.

Thanks again.

[ii00907]

I William,

your change on rescanAllHBAInCluster.pl, for my request, is good.

I have insert on my "Production Process" and i have reduce ( 15 Minutes ) the process for clone e mount DB-rdm.

Many thanks