esx5i + VSA HP P4000 Boot problem.

[Florian Schmutz]

Hm u can schedule this over multiple points and also u can do this in many different ways...
on example u use windows Scheduler and create an script over powershell oder vcli.....
One other option could be to use a cronscript on esxi hypervisor as self.... but any way its like i posted... i dont engeneer this but there are messi ways.... sorry for that

[5y53ng]

---

@doublejae wrote:

Anyone come up with a solution to this yet??  In my mind this is a HUGE issue with the VSA solution and I can't believe that is hasn't been addressd by HP!!

I've heard that there is a script solution to force the ESXi host to perform an iSCSI rescan however the issue becomes timing.  How do you schedule this script to run AFTER the VSA VM boots and BEFORE the rest of the VMs try to power on??

---

vmware HA will take care of restarting the VSA after a power outage. The key to knowing when to rescan is to check and see if the VIP is available. If you can reach the VIP then you can rescan your HBA.

This script has to be planned carefully, as their are conditions when you would not want to force a rescan. The following conditions should have your script exit...

1. In maintenance mode

2. iSCSI Initiator is disabled

3. Your iSCSI disks are already mounted

If above conditions are false, then you want to see if the VIP is available. Checking for the VIP should only loop few times before giving up , since it is possible the VIP never comes up due to other problems. Wait 15 seconds or so between pings to the VIP.

If the VIP is reachable, then you would want to force a rescan, but again this should only loop a few times in the event the targets are never discovered. After rescanning check to see if the disks are mounted again.

You'll need average scripting skills to write this.

[ialyoshin · Accepted solution]

Probably outdated already, but you can use /etc/rc.local in ESXi to do a rescan after VSA startup. Example script is here - http://it-memories.info/?p=15 

It's quite dirty, but you can improve it.

[Florian Schmutz]

Nope not outdated. I Will open a internal case to check this but its the first Reply wich sounds very helpful.

Thanks for your Post

[5y53ng]

If you add the following checks you essentially have the same script I am using;

1. Check for host in maintenance mode

vim-cmd hostsvc/hostsummary | grep inMaintenanceMode

2. Check for iSCSI enabled

esxcli storage core adapater list

3. Check if iSCSI disks are mounted

ls –l /dev/disks | grep naa.6000eb

4. Check to see if VIP is reachable

ping x.x.x.x

Add some counters so the script gives up after a predetermined number of loops. Also, log the output so you can go check to see what happened after problems.

Unfortunately I cannot post my script here. I aksed for permission and was refused.