First off, congratulations to Brent, Eric, and everyone at Whamcloud. I had thought that the BI/Big Data side of things could prove interesting for them, and might make them in play. I hadn’t realized how quickly this was the case.
Second, Big Data is huge. Lustre, which is effectively Whamcloud’s product (ignoring IP ownership, yadda yadda …), can play there, though it needs some serious additional work. But with the acquisition, I’d argue that the multithreading MDS and ODS are not far off. Mebbe even supporting some non-ext4 file systems (cough cough … xfs, btrfs, … cough cough)
Honestly, I had thought Netapp or EMC was going to do this. I hadn’t counted on Intel. This does make sense though, given what I am seeing in the higher end of the market. Intel doesn’t want to be a chip company. Its competitors are full on systems companies, with switches, motherboards, chips, OSes. …
Intel bought Qlogic’s IB product lines. It makes sense that they buy one of the things that IB makes better.
I expect more consolidation (looks around for lawyerly looking guys with briefcases). More acquisitions, etc. Companies like Instrumental and alike could be snapped up.
Two years ago, I didn’t think Lustre had a future. I thought the Sun/Oracle HPC divestment would largely render it a fragmented environment … with 3 different organizations … with different goals, serving different constituencies. There were alternatives, and they are good.
One year ago, I was guardedly optimistic. Wasn’t sure if there was a long future, but people were moving in the right direction. Again, alternatives, that are good.
Today, with Intel purchasing them, this is a significant boost to its likely success. Intel buying a company is very different than, say, HP buying a company. The process is additive in the Intel sense … they are there, they have more resources.
This is good. Lustre isn’t going away now. And also, its likely that some of its long standing issues in implementation will be fixed. Specifically those dealing with multi-threading and NUMA. Intel has … at least a little … interest in this? Mebbe?
If you look at the Intel stack, about the only thing they are missing is the OS.
I’d wager that Red Hat is being looked at.
And no, I have no inside information whatsoever. I am probably badly wrong (and I can see some private emails headed my way over this). But that one makes perfect sense. That one would be a no brainer.
And for laughs, I’d also argue that ScaleMP is very much in play. Take smaller units and aggregate them together to build larger units. Make that part of your OS. Provide both cloud/hadoop file system (gluster), HPC/big data file system (Lustre). Allow machines to be disaggregated (KVM) and aggregated (ScaleMP vSMP) on demand. All you need is an integrated scheduler (PBS/SGE/Slurm/…).
Yeah. It makes perfect sense.
Allow me to repeat: I have no inside information whatsoever.
It will get interesting when someone buys Seagate and someone else buys WD. Yeah … that gets real interesting then.