Lefthand 12.0

[vlho]

Hi all,

HP StoreVirtual Storage LeftHand OS 12.0 is out:

https://h20392.www2.hp.com/portal/swdepot/displayProductInfo.do?productNumber=StoreVirtualSW

[Prakash Singh_1]

Hi,

The link refers to Left Hand OS 11.5....There is no such release as such. Kindly Share appropriate link if possible.

[Davy Priem]

Thx for the notice. Not everything is online though. CMC version 12.0 doesn't download the upgrade files yet and the site containing information about memory in a P4000 G2 system (www.hp.com/go/P4000G2memory) isn't online either.

[psmiano]

It's on the FTP site as of 2/21. Not finding any documentation....yet.

[Bart_Heungens]

Yes it is there (Software Depot - Storage softwxare...)

Have installed it already, all my findings and details (with screenshots) can be found on my blog

http://www.bitcon.be/?p=2904

Kr,

Bart

[a_o]

Anyone else try to upgrade an existing cluster with 12.0?

Ran update through CMC on a v.11.5  mult-site cluster.

I now have a borked cluster.

Here are the juicy bits..

Ran CMC update procedure on the cluster, and accepted the prompt about the client side updates.

FOM upgraded fine, after bumping up the VM memory.

I'm not too keen on the change to  hiding the Linux boot process .

One storage node upgraded fine and showed a 12.0 console prompt after about 30 minutes.

But the CMC could not contact this node, and promptly halted the upgrade process.

The upgraded node could respond to pings, so I tried firing up another CMC session, and still could not connect.

The other node did not complete the upgrade process, and is still at 11.5.

Eventually, I had to reset the non repsonsive node, as it just sat there with Hyper-V CPU utilization of 4%. 

Seemingly, it restarted. This time, there was no SAN/iQ prompt. It sat there for over 12 hours like this.
I was thinking maybe it was doing a linux disk/journal check or something.

Finally, I had to cancel the remainder of the  update process in the original instance of CMC.

Before I rebuild the cluster by re-installing the VSA nodes, does anyone know of a way to enable the showing of the Linux console boot process?

Good thing, this cluster was used as  secondary target for my Veeam backups...

[oikjn]

I upgraded a 2-node AO 11.5 group to 12 without issue, but there is little on this group.  No problems seen so far for me but it isn't really doing anything yet.  I didn't see any unmap recovery yet.  I"ll have to read up about that implimentation.

[Andre van den Berg_1]

Yesterday installed 12.0 on 6 node cluster with 2 TIER on multi-sites. First updated all the host in the Hyper-V cluster with DSM en Snapshot software and then upgraded to 12.0.

Everything went smooth, now even SMI-S is working so can see the storage in System Center Virtual Machine Manager 2012 R2 with CU5.

[thae]

Hi,

can anyone provide with the MEM offline bundle (not the vib)?

According to the StoreVirtual Multipathing User Guide, it should be available at the HP vibsdepot but i can't find it anywhere...

Thanks,

thae

[Bart_Heungens]

Looking for the file myself as well, does not seem to be published yet on the vibsdepot site... Should be there soon...

I did the trick with the VIB file from the Software Depot site, works fine... Writing an article on it at this moment on my blog...

Kr,

Bart

[fourg]

I submitted a support ticket in hopes to get the offline bundle, I also can't find it anywhere.

Update: HP Support acknowledged the offline bundles were missing.  I was sent the one for vSphere 5.5 which I was seeking and I've attached it if others want it.  I was told I'd be contacted when they got the other offline bundles online.

Update 2: The attached offline bundle is actually for ESXi 5.5 even though the filenaming says 5.5.  Waiting on HP to provide the true 5.5 bundle.

Update 3: I've been asked to remove the attached offline bundle zip file by HP Support.  I was accidentally sent an unsigned copy and have been told I will need to wait for HP to get the bundles properly signed by VMware.

[thae]

Great, any feedback yet from HP support?

I'll open a ticket as well otherwise.....

[thae]

Thanks a lot for posting the offline bundle.

Though named 5.5, this one seems to be for ESXi 5.0?

Did you receive 5.5 as well?

[fourg]

I encountered the same issue with it actually being for ESXi 5.0 even though the filenaming said it was 5.5.  I've been told the ticket has be escalated to L3, so hopefully sometime soon they resolve this.  Concerning that I still dont see these offline bundles anywhere to be found online either.

[Leslie Thomson]

The bundle in the thread here should not be used because it was an unsigned copy and HP will soon make available the signed copies via software depot. This will take a minimum of 2 weeks because the signing process is out of HP's control.

[s_mcleod]

Upgrader beware!

The upgrade process is incredibly painful, many reboots of the storage systems required, java crashes etc...

Book out at least 4 hours per cabinet and some kleenex.

[oikjn]

not sure about your environment, but this upgrade is no different than any of the others.  only catch might be the need to increase memory in the FOM or on a VSA, but that isn't unheard of either and really minor.  The time to upgrade depends on the environment size and load, but each node upgrades one at a time and reboots one at a time which means there is a restripe for that one node each time which can mean that the upgrade process will be very slow if your SAN is under heavy load and the amount of data that changed during that node's upgrade phase was significant.  99% of the time, the upgrade process itself is quick, but its the resync after which is the real killer, but this is no different than any other update on this SAN.

[Rascii]

I'm not finding any memory requirements for the FOM. Can anyone comment on what they actually are? Thx.

[thae]

Any news wrt to the MEM offline bundle?