by
adam baum
Day two
My first session of the day was on the topic of VM in gov't and education. The big takeway on this one is collaboration. The panel members' project were very successful because they collaborated with other gov't (or educational) entities. A great example is DR. Two counties got together to act as each other's DR site. Vm is a great enabler of this since the hardware outlay is small compared to the cost of a full data center. There is a sessions about two univerisities doing exactly this. One is on the East coast, the other on the West.
Each of the panel members discussed how VM came to be used in their organization and it's benefits to them. I am guessing that these reasons and benefits are just universal truths. What does that mean? Think about VM selling/maerketing points: Consolidation, efficient utilization of resources, less real estate needed, etc. We all listened and now we have VM. I'm sure there are a few companies out there who are running vm just so that they can say they are running it, but I doubt there are very many.
The session on virtualizing Exchange 2007 was so-so. The presenters really just went over a bunch of sildes detailing some tests performed and results. OK. But what did they learn? Not a whole lot other than that it's very doable. One statement sort of made me laugh. The results show that if you have a large number (2000 or more) of mailboxes on one server, don't share the host with other VMs. Heck, if I can't share the host why virtualize? That wasn't the point of this session. The session was about "Could it be done?". So the answer is yes, but don't have large mailbox servers. I may need to revisit this post in a week or so after the session materials are posted. I'm wondering if I misinterpreted what the presenters said.
For my third sessions of the day, I went to see how Health South uses vRanger (formerly known as esxRanger). Health South teamed up with VizionCore for this one. Note to presenters: you should meet and discuss your presentation ahead of time. These two people never met until this session. Each had a portion of the presentation and they really didn't go together. Health South presented its case and extolled the virtues of vRanger. VizionCore provided some useful info. Instead of talking about the product, the presenter talked about how VizionCore does backup at its shop. They break their backup methodolgy down into DR requirements. Those considered tier1 get backed up and replicated to the DR site multiple times throughout the day. Tier2 maybe two or three times a day and tier3 gets a traditional backup nightly.
My last session was Advanced 3.x Diagnostics Logs Analysis. Ha! The presenter opened up by saying that this is usually a 2day session. He went on to clarify that the materials presented were what the tech support folks at Vmware get when they are trained. I'm thinking this is going to be real good. Ha!. It was good, but it was way too fast. Not only too fast, but a bit over my head. The session title does have the word "Advanced" in it. Next time I will know better. Anyway, Mostofa (the presenter) listed where the bulk of the log files can be found and what they contain. He then proceeded to dissect one. In this case, we watched him dissect a LUN discovery and a SCSI Bus Reset. You know things are going to be a bit difficult when they tell you to download a sourcecode file and look in one of the header files so you can see what the result code means. The header file has a comments section that lists what each result code means. BTW, since we are working in the service console the sourcecode is opensource. I did come out more knowledgeable. I was just overwhelmed. Here's a takeaway: SCSI Reservation Conflicts are OK. SCSI Reservation failures are a problem. Now you know.
End of day, time for food and sleep. I skipped the big party.
(Note: You must be logged in to post a comment.)