Just waiting for the Russians to blame US radar or anyone other than themselves.
26 posts • joined 9 Apr 2012
Can we leave the religion out of it?
The Hadoop environment just feels so unpolished, even using one of the "supported" (cough) distributions. Getting even simple queries to work right felt like building some kind of Rube Goldberg/Heath Robinson machine. No wonder there's a lack of enthusiasm from anyone other than the cool chasers.
I always assumed that Glass was deliberately clownish to distract attention from Google's data mining operation. I suppose the same goes for Eric Schmidt too.
Cisco would do a Flexpod deal with a dog food vendor if they thought they could move another UCS and switches.
Since HP isn't merging with EMC, they're now trying to copy its Federation? Love it! Apparently The Machine which secretly runs HP (and Meg like an audioanimatron) still has a few bugs.
They're _all_ paid for reports. Some are just less obvious than others.
Re: No surprise
You need to put down the crack pipe. Nimble isn't anything like NetApp internally or otherwise. Nimble is a purely iSCSI product based on a log structured filesystem with a large SSD cache in front of it. You obviously have the same deep knowledge of NetApp that you have of Nimble.
Re: Oracle new ZFS servers
As Hadoop is HPC for morons, and there's lots of morons, there's lots of profit (or that's the story, anyway).
Given how well connected the founders of this one are, they could have announced they were developing chocolate dog food and would have still been bought out for the same premium. Once you've achieved the appropriate level of connectivity, one barely has to go through the motions of innovation any more.
I'm thinking of starting my own storage service involving encoding data steganographically in pictures of Justin Bieber uploaded to Facebook. As crap is unusually persistent, the required level of redundancy is low, plus FB will be picking up the tab. Voila, efficient cloud storage! Anyone in Cisco or EMC bizdev interested?
Re: Mixed feelings from an American user of Huawei devices...
My take is that there's a pantheon of evil.
Apple just wants your money and at their margins doesn't have to plot that hard for it.
Microsoft wants your money too, but at commodity margins they have to be everywhere hence their deeds.
Google doesn't care about your money, only about selling you as the commodity. The more they get on you, the more money you're worth to them. It's that objectification/dehumanization that makes them extra special evil. With bonus points for extreme hypocrisy along the way.
Oracle are Butlin's Redcoats by comparison.
Seeing the luckiest man in the world try to bask in Assange's glow is grotesque. Schmidt should be hanging out with Wozniak sharing tips on seeming relevant.
I hope they get a vote at the papal conclave out of this.
If they're Software-Defined Storage, then so is every Linux ISO image out there. Heck, what's Windows Storage Server 2012 if not this?
The memristor will reduce them to rubble! Jeez, can't you find something more plausible, like Woz inventing anti-gravity at Fusion-io?
Maybe Cisco should buy or merge with EMC instead?
Minor HP Nit
Minor nit re HP "one with a paralysed mid-range due to perceived EVA end-of-life and no 3PAR replacement products". The recent StoreServ (awful name) announcements have remedied this well, even if they'll probably "rightsource" development and support to Kreplachistan to save a few bucks.
3PAR File Services?
Is File Services integral to the 7000 series or is it just a separate Windows Storage Server server?
Re: HP Violin
Exsqueeze me, Meg Mountain: 3PAR thin provisioning definitely introduces performance variability*. Effectively you're converting sequential I/O into random I/O as you hunt around the disks for wherever your many 16KB blocks actually reside. It's a significant performance hit in real life.
3PAR thin provisioning _hype_ may be light years ahead of everyone else's but they're all serving the same lumpy porridge.
*Don't even get me started on 3PAR non-deterministic latency issues that I've experienced.
Stable revenue is hardly "flat-lining"; that would be HP. Were you out of cheesy sex metaphors and had to settle for death instead?
Aren't CTOs at NetApp like VPs at banks?
Jeez, another mention of "memristors". This should be part of the Reg drinking game along with penile metaphors.
Your choice with Gluster is either to use a server node with an NFS client or via the FUSE client. NFS is more efficient in the sense that the client is better plumbed in to the system, yet even with FUSE overhead, the other is a cleaner mapping from Gluster to local filesystem. As the article itself states, the FUSE client has "some performance benefits". I used it mostly because it was less of a pain than the NFS functionality.
FYI, the "Fuse" client isn't a native client per se. Fuse is mechanism for letting a user-space (e.g. non-kernel) program appear to be a filesystem and as such, incurs more overhead than something like NFS which is a kernel module.
Gluster's weak point, which is not mentioned, is that a file cannot span server nodes. If you have 50G left on one and 50G left on another and need to write/copy/save 51G of data, you're SOL.
Last but not least, "Shadowman"? What dark and odorous orifice did you pull that one from?