I was invited to Netapp insight 2010 in December and I didn’t get a chance to write much about that at the time being end of year rush and all. So here are some thoughts. IBM N Series and Netapp are essentially the same thing so I will use either term generically. I’m not mentioning presenters names as I’m not 100% sure the names on the slide decks I have were the actual presenters in Macau. Disclosure here is that not only do I work for IBM, but Netapp funded my trip to Macau. Even though I booked my 3 days solid there were still a lot of good sessions I couldn’t get to because of conflicts.
The elective sessions I attended in order were:
CommVault: VM Protection with CommVault and Data ONTAP Snapshot Copies
- The session was OK, but the main value I got was in realising that the presenter was very knowledgable and so I took the opportunity to have a chat with him afterwards. One thing I wanted to clarify was Commvault’s preference of CIFS targets for d2d backups so as to avoid LUN locking issues when there are multiple backup servers.
Oracle Internals: The Basics
- I heard some people say they weren’t going to this because it might be too ‘basic’, but the presenter was talking about INTERNALS, and he was very knowledgeable. Even the basics about the internals of ASM and RAC can be quite complicated and this session was pretty interesting and included tips and myth-busting e.g. hot backup mode does not mean writes are only to the logs.
Storage Efficiency in Action
- This mainly focused on the zero fat, low fat and full fat provisioning models. If you thin provision you must monitor capacity usage carefully and plan your responses to space-low conditions. Netapp recommend the low fat model for most situations and involves fully provisioned LUNs with thin provisioned snap space.
NetApp Data Replication Deep-Dive
- This was a big session which covered SnapMirror, SnapVault and OSSV (host-based SnapVault) including interaction with dedup, differences between ONTAP versions, and also touched on comparison with MetroCluster. Lots of good detail, but perhaps a bit too much time spent on OSSV in relation to its market uptake.
Storage Subsystem Best Practices: Configuration and Resiliency
- Again lots of good detail from a knowledgeable presenter e.g. we heard that in general Netapp is now recommending larger RAID groups (up to 18+2, or 26+2 in the case of SSDs). We covered rebuilds and how the 24 drive SAS cages have proven far more reliable in the field than the FC cages they have replaced. We also learned that Netapp is moving to use full binary capacity counting rather than the unfortunate fusion of decimal and binary they have traditionally employed.
Sizing for SnapMirror (Disaster Recovery) with Real Application Workloads
- This was really a click-through of the tools available under Synergy for sizing SnapMirror. Click-throughs don’t work very well as presentation content IMHO and this ether needed to be a lab, or else more focused on the principles and less on the clicks.
Selling Solid State Disk and Flash Cache
- The essence of this was that Netapp will offer full-trays of 100GB SLC SSDs (interesting choice), but they believe that most customers will derive much better value from Flash Cache (PAM II) than from an aggr full of SSDs.
Thin Provisioning for the Impatient
- This was a more detailed look at the zero fat, low fat and full fat provisioning models with cautions and management recommendations to get best value with least risk.
Antivirus Scanning for Windows File Services
- I just wanted to check my assumptions about how this was put together and things to watch out for, and it was pretty much as I expected. Interesting to note that Netapp are planning to support on-controller anti-virus for NFS and CIFS at some stage (McAfee & Sophos initially) which will eliminate the network latency associated with offloaded scanning.
Architecting a NetApp Data Compression Solution
- This was a really good session. Netapp have done a great job of bringing dedup and compression into mainstream usage at zero license charge. There are however a lot of small things to be aware of, such as interactions/limitations around snaps, snapmirror, flash cache, data motion etc. The Netapp recommendation was not to use compression where performance is critical. The following chart of use cases might be helpful.
VMware and NetApp Integration: Today, Tomorrow, and Beyond
- This session was a bit too long and fluffy and repetitive for me. Looking back over the slides I can see there was some good content there, but at the time it felt like death by Powerpoint. Maybe partly because it was my last session : )
Slip of the event goes to the Chinese concierge who when I asked him if there was any likelihood of Macau adopting the Hong Kong dollar (Macau pataca is pinned to HK$ anyway) replied “It’s not possible – Macau is an entirely different country!” At which point I reminded him of China’s “one country, two systems” slogan. The poor chap has probably already been hauled off in chains.
And to sum up Macau as a place, I’ll leave you with this photo of St Paul’s facade built in 1640, surrounded by plastic Christmas pandas.
Filed under: N Series |