

First off, special thanks to Jason Boche for finding the root cause of this issue. Turns out in an OVA deployment the time syncing with ESX and VMware tools will cause problems with the SSL cert generation. Basically with some setups it will bounce time 5 hours forward *before* the SSL cert is generated. This is related to some time zone funkiness I have a fix for. Problem is this means that cert is always invalid because it comes from the future where Chad Sakac is fully cyborg and Microsoft just makes wallpaper. I will build a fix into the future versions. But, if you want to fix now follow this simple step: Login to Celerra UBER VSA as root. Default user/password is ‘root/nasadmin’. Make sure time on VSA is synced. Use NTP or set manually. Run the following command: /nas/http/webui/bin/gen_ssl_cert.pl This will regenerate the SSL certificate and should fix any blue screen issues. Thanks, .nick


3 weeks is a long time to work on something. I think that is longer than any Celerra VSA revision, tool, or project I have worked on since EMCWorld. It is funny really. Doing this video helped me find two different bugs with the Celerra UBER VSA (see here & here) and a discover a bug with the Celerra Replicator SRA (here). The results are though are worth it. Below you will find a video on how to build a complete UBER VSA 3.2 / SRM 4.1 lab using VMware Workstation. This tutorial will walk you through the install of the VSA’s, SRM, plug-ins, networking, storage configuration, mounting to ESXi, configuring SRM with the Celerra Replicator SRA, and finally testing and running a live failover. This includes both NFS and iSCSI configurations and the final test shows you how to setup SRM to failover both NFS and iSCSI datastores within the same configuration. I am not kidding when I say this was a lot of work. And I have taken a different approach to making a technical tutorial video than most. My goal was to make this somewhat interesting and as short as possible given the amount of config. Hopefully I will make you crack a smile. Even with cutting out loading screens and how fast I talk, this still weighs in at a pretty long length (1 hour 13+ min.). I am pretty sure I can build this in my sleep now and as great as both products are, [...]
It came to my attention pretty quick that something was amiss with the last Celerra VSA UBER release. I heard strange stories of disks not adding and the OVA not deploying. Since this thing is drawing close on a couple thousand downloads some environmental, transfer, and operational bugs/errors will cause problems. But I finally nailed this down myself when I was doing my iSCSI testing. I had run into a SRM bug (see post here) this week. And as I was parsing the logs I noticed a couple lines that confirmed it. There was definitely a bug as iSCSI objects were not pulling the new instantiated ID that is created with the wizard. So after 8 hours or so of diving through and reversing engineering I finally found the culprit. It went all the way back to the original build I get from engineering. I wrote a patch, tested against my running VSA’s, and confirmed I had the fix. But while I was about rebuilding the VSA’s I decided to do a couple more things that didn’t make the last list. Here is the running list of changes: Bugfix: Passphrase for peer connections is will now save correctly. This is related to the ID bug. Before if you rebooted the VSA replications (NFS or iSCSI) would no longer work. Bugfix: iSCSI replication now creates LUN’s and Replication sessions with proper naming ID’s. Bugfix: NTP settings for Data Mover will update time during wizard immediately to correct large skew. Bugfix: Either [...]


So transparency is a good thing right? Especially since I just found out I have a ‘vendor’ blog *eek* I will make this short and sweet. I have been working on a home lab SRM thingy for a couple weeks. I had completed all my testing with NFS replication and plug-ins. Pretty darn sweet stuff but that is another post coming. When I moved on to the the iSCSI testing everything was kosher until I ran into issues with the Celerra Replicator SRA. It will crash out with a ‘discoverLuns’ error when searching the Celerra for the replicated session. In my lab I am using the new DART 6.0 UBER VSA. Turns out this issue is known and just related to a naming change. EMC already has a fix (I tested and it works) and it is on its way through the official channels as we speak. So once again with iSCSI replicated sessions on DART 6.0 there is an issue with the ‘discoverLuns’ command only. This does not affect the NFS replication with the Celerra SRA or other Clariion, Recoverpoint, Symmetrix replications/SRA’s. As soon as the release is out watch for it here or on Chad’s site. .nick


Man this has been a busy couple months. I have a whole different blog post about VMworld and the MADNESS. But suffice to say I have been busy. So busy in fact that I didn’t really have time to work on the new Celerra UBER VSA until this week. Which also happens to be a week vacation for me. Now for a normal 9-5 kind of guy, working a little bit on your vacation is ok. Some people frown, some people smile and approve. But, on the vSpecialists this is a big "no-no". So because of my strong desire to get this awesome piece of virtual goodness in your hands I have risked the smack down from my teammates and managers for working on it during my vacation. I shall likely be banned from a console for a bit so remember me when you are checking out the UBER cool new Unisphere interface. And so I am excited to announce the release of the Celerra UBER VSA version 3. I got quite a bit of new features as well as the new code/management interface you have been hearing about. Here is the list of changes and additions: DART is now 6.0.36.4 Unisphere management console (rocks!) The Celerra VSA is now 64 bit! This means you can throw RAM at it for bigger setups and it will use it. Over 8GB becomes less beneficial without code changes to simulation services. Future updates will fix this from the Celerra VSA engineering teams. [...]