I know of a few others that have similar setups as mine that had issues upgrading to OneView 3.0. When we attempted the upgrade from 2.x to 3.0 it ereased our FC config, disconnecting all of our hosts from our 3par storage.
Our setup had OV managing the enclosure it was living in (bad idea I know) we've since fixed that by moving OV and a few other things off to a dedicated host with local storage.
Has anyone else experianced an outage while performing the upgrade? We ended up spinning up a new OV2.x VM on local storage, restoring the backup. Once we did that and rebooted each host we were able to bring up the VMs.
I experienced something similar. In fact I just finished getting our lab back online....ish
I also had a OneView living inside the environment it was managing (what can I say, it's a lab so its the wild west) and at some point during the upgrade it looks like all the the storage paths tanked. We have our storage connected to a Cisco MDS/Isilon/VNX. This of course brought our lab to a screeching halt because all of Hosts lost communication with the VMs. We ended up bringing another ESXi host online on a rack mount, and zoning it into the same LUNs we had our VMs on, then we cherry picked out the neccessary ones to get our lab back online, and so that I could restore our OneView VM back to vers 2.x
Needless to say, I have experienced the same type of issue. Has anyone opened up a support case with HPE so that they can look into this? Once I get my lab back to 100% so that I can remotely access it, and work with HPE I will open one, unless of course we get a notification of what is causing this issue/workaround.
I have a case with HPe support already open. I was lucky enough to be the first customer to call support with this issue. I don't have a lab environment that I can try this in again but have a partner who is trying to get it working.
HPe released an update last night. IF you did successfully you will probably want to upgrade to 3.0, update to 3.00.05 asap. From reading the advisory it sounds ilke just a simple reboot will cause the FC config to be removed.
The advisory pasted below will also show that doing an e-fuse reset of each affected blade will resolve the issue.
The CA @sheath7288 posted outlines the upgrade issue customers unfortunately will experience IF they proceed forward with upgrading any prior release to 3.00.04 AND use Virtual Connect FlexFabric or Virtual Connect FlexFabric 20/40 F8 modules with Fibre Channel uplinks. We quickly found and issolated the issue, and re-released HPE OneView 3.00 with a new version, 3.00.05.
For those customers who have already upgraded to 3.00.04 and do not have the issue described, or who already have upgraded and do not want to downgrade to the prior version, please download the 3.00.05 Update bin file.
Please, follow the CA, and any updates to the CA will be made within it, not here in the forums. To be notified of future CA, please use the Proactive Updates Subscription Form to subscribe to your relevant products. You can use either your SAID contract to enable alerts for your products, or use the Simple Search to look for HPE OneView and add to list.
