Texas Memory Systems (TMS) is getting some enterprise cred – by introducing a high-availability, shared-access RamSan flash array. Competitors such as Violin Memory have criticised TMS RamSan flash SAN arrays for not having high-availability (HA) features that enterprises require when running business-critical applications. TMS …
I've no doubt it will be amazing
They already had street cred in my mind - for years they've been churning out RAM based arrays that drop excrement from a height on other storage solutions, as far as speed is concerned. Never used them myself, but I've known about them for a long time.
hot swap, online upgrade ?
are those redundant components hot swappable ? are there online software upgrades on the box or are those things you need the 'application high availability' for -- if so then there's no new street cred here.
I too, like the look of this beastie
I first had my attention drawn to TMS when they were advertising 1,000,000 IOPS for their RAM SAN (that has now been clipped, probably by the marketing mavens or more truth in advertising).
Sho' enough ..... at these prices flash array may soon become a credible altenative to traditional FC, SATA or SAS disk arrays. So please do rock on... Our heavily discounted FC 4 G/B SAN disks still cost us about 1200 £ a TB a couple of years back. But performance? As usual, it depends on how you crunch the numbers. FC can provide a sustained data burst of up to 700,000 IOPS on something like the DS5000 but I'm not too sure how that raw measurement comparably applies to a flash array. The flash vendors may be using average (or constant) rates over a sustained test to come up with a more realistic figure which takes everything into account (like no platters or disk heads). I still haven't seen a decent and honest race comparison between any of these SSD offerings and current "gold-standard" Fiber Channel products.
For those who could only afford SAS or SATA disk arrays, however, the new SSD products will pound them into the proverbial performance dust (which is why SATA and SAS are nearly always used for comparison, methinks). But then the price differential yawns all the more hugely.
The real sex appeal of a flash SAN array comes from its lower power consumption, faster access, and higher reliability (not to mention the reduced space requirements). But is all that worth paying at least 10x the price of today's high end storage products per TB? Maybe, wait and see.....
If a flash array SAN (I think we will soon need a new acronym here, how about a FASAN?) can do what our big-pig, dually-redundant monolithic FC disk structure does today, better and faster (and more affordably? ...oops) we probably wouldn't buy any new FC storage. But that is still a big IF right now. We won't because we must first amortize an initial large investment. SANs have a comforting tendency (for the buyer at least) to stay in production for a long time ( > 5 years). Our last storage array served reliably for nearly 10 years. Hardware prices tend to become much lower when working within such extended time frames. However, for the budding or established cloud vendor who needs to replace a creaky storage infrastructure, now may almost be the time to go flash (depending on their applications and budget, obviously).
Anyway it is all VERY exciting to watch. But I am glad our shop doesn't need to decide now.
Cool, but I think they are about two years too late. It's crazy that an "enterprise" level storage vendor didn't have a SPOF solution before this. Well, at least with this they won't have to worry about bringing down any more Australian airlines.
I think it was a bad architecture that bought down the Australian airline,
- Product round-up Ten excellent FREE PC apps to brighten your Windows
- Review Tough Banana Pi: a Raspberry Pi for colour-blind diehards
- Product round-up Ten Mac freeware apps for your new Apple baby
- Analysis Pity the poor Windows developer: The tools for desktop development are in disarray
- Chromecast video on UK, Euro TVs hertz so badly it makes us judder – but Google 'won't fix'