|
evil_bunnY posted:Equallogic doesn't offer SATA at all, even their 7k drives are SAS. I personally replaced several SATA drives in my PS6510E; all my 48 drives are the same 2TB 7200RPM SATA-II Seagate Constellation, like this one here: http://compare.ebay.com/like/261063571150?var=lv<yp=AllFixedPriceItemTypes&var=sbar
|
# ¿ Jul 31, 2012 15:49 |
|
|
# ¿ Apr 25, 2024 05:01 |
|
Nukelear v.2 posted:Just as a follow-up on this, the host I was testing from did not have it's nics set for jumbo frames. Changing the MTU to 9000 resolved the issue, not exactly sure why this would have happened though, should have just been performance degradation as far as I understand things. It's now happily pushing close to 33,000 IO/s at 255MB/s on my 8k random workload. Do a search in my posts about Broadcom and you will see why all their drivers are utter PoS.
|
# ¿ Jul 31, 2012 16:18 |
|
Misogynist posted:They're useful as giant scratch storage, but my personal recommendation is to never store anything important on DDN. I used to run our entire production load on a single S2A9550, there's nothing wrong with them, I think. It's just that it's a big, dumb RAID storage piece, there some options - strictly console-only mgmt - but basically it was a giant fast array (~20TB) and that's it. No features, not even snapshots, nothing. szlevi fucked around with this message at 20:17 on Jul 31, 2012 |
# ¿ Jul 31, 2012 16:33 |
|
Vanilla posted:So they're not even on the NAS one as it doesn't do NAS right out the box. If you do a google search for the Quadrant you'll see it's EMC and Netapp in the top right (2011 is the only one I could find online). You are behind the curve - there is a dedicated EqualLogic NAS called FS7500 and it's a pretty screamin' good box, running Dell's Fluid FS which is built on Exanet's file system (Dell bought Exanet's IP in early 2010, for Michael Dell's ashtray change.) The nicest thing is EQL's signature simplicity: it integrates into the same browser-based mgmt stack, you can manage any EQL unit via any other EQL unit, SAN or NAS.
|
# ¿ Aug 16, 2012 18:49 |
|
Powdered Toast Man posted:Boss is leaning towards getting some Dell/Equallogic poo poo because I guess they're in the far upper right of Gartner's magic quadrant for SAN/NAS vendors. Is the stuff any good? I have yet to meet an EQL user who wouldn't praise his box - it's fully redundant inside, very easy/simple to manage, does deliver the performance and with the upcoming new firmware release it will do pretty much everything you can ask for incl. synchronous replication, re-thinning, undelete etc etc.
|
# ¿ Aug 16, 2012 19:55 |
|
Wicaeed posted:Unfortunately that's the one thing we are lacking (spare networking hardware) Nevertheless setting up your EQL shouldn't take more than 30 minutes (worst case.)
|
# ¿ Aug 16, 2012 19:56 |
|
three posted:You have to configure all of your hosts to tolerate ~40 seconds of the EQL array being unavailable during firmware upgrades. Maybe this is recommended with other iSCSI arrays, but I haven't ran into it yet? I never did it but IIRC my values are ~30 secs and my hosts all tolerate failovers just fine...
|
# ¿ Aug 17, 2012 01:22 |
|
Wicaeed posted:Does anyone know much about the Dell Online courses for Into to EqualLogic PS Series Storage Arrays? Specificially if we purchase training for them (but share a login for the account the training was purchased on) can we view the training at any time even after completing it? I cannot fathom what they can teach you that you cannot learn yourself in a few days, for free...
|
# ¿ Aug 17, 2012 01:22 |
|
NippleFloss posted:iSCSI is much easier for most general purpose IT people to grasp. Fabrics and zoning aren't too tricky but there is a learning curve there. Additionally when you get into FC you're also getting into the business of ensuring that you've got solid HBA firmware, that you understand the vendor specific MPIO suite you're using, that you understand the OS specific tools provided to manage those HBAs. And it's still much less likely that you have anyone on staff who knows enough about FC at the protocol layer to troubleshoot difficult issues, while it's quite easy to find IP expertise. Yes + iSCSI is a lot cheaper, even in 10GbE flavor: show me a 24-port line-rate FC16 switch for $5-6k... ...did I mention that for IB you can get a 36-port FDR switch for ~$8k?
|
# ¿ Aug 17, 2012 23:34 |
|
madsushi posted:NetApp's various host utilities (SnapDrive, VSC) will set these timeout values for you automatically. Yeah, that's possible, that EQL's HIT sets it at install...
|
# ¿ Aug 20, 2012 17:05 |
|
KS posted:Compellent arrays do all writes to 10 and rewrite to RAID-5 in the background. Last December I was told it's RAID10 and RAID6...
|
# ¿ Aug 20, 2012 17:06 |
|
Misogynist posted:The Windows support for NFS serving is incredibly poo poo, though, and good luck doing access control unless you're doing a full Centrify implementation. THIS. And I'd add that unless you go with Server 2012 and all it brings to the table - eg not being able to run ANY 2012 RSAT tool on Windows 7 - your SMB/CIFS speed will be also a pile of poo poo compared to a dedicated/non-Windows unit.
|
# ¿ Nov 21, 2012 19:13 |
|
three posted:I sure hope you have a fast link if you're using replication as your backup. Semi-related: people using replication are saying the new Equallogic 6.0 firmware (the one that brought us synchronous replication) have indeed fixed the bandwidth bug and they are seeing 100Mb/1Gb links saturated... anyone here can confirm this?
|
# ¿ Nov 21, 2012 19:17 |
|
Interesting rumor: http://www.theregister.co.uk/2012/11/19/dell_afa/ Thoughts? Despite my reservation about Nimbus (see my comments there) I'd say it is still far the best price/IP ratio for Dell - question is if they want to commit enough money to make it fully integrated into their storage product line (a.k.a. enterprise-ready)...
|
# ¿ Nov 26, 2012 22:51 |
|
NippleFloss posted:I'm not really sold on the need for an all-flash array for 99.9% of consolidated storage customers. Is anyone using this? For what? What workloads out there do you have that need 1 million random read IOPs with a total capacity of like 5TB, at a price multiple of many times more than a traditional storage array with 10 times the capacity? madsushi posted:I have been asking myself the same question. I think that your primary use-case is going to be someone with a relatively small but intensely used database that's too big to fit in a FlashCache card (multiple TB). Our workload - high-end medical visualizations, compositing, working with 4K and higher datasets, volume rendering etc - is a classic example... frame sizes start around 16MB and regularly reach 128MB or even higher and I need at least 10-15 fps (over ~1GB/s) per workstation so that's at least one of Nimbus' boxes IF it is indeed able to deliver 10-12GB/box - latter needs to be tested, we are talking about the shittiest protocol (CIFS/SMB2.x), RL is always worse than specs on paper. That being said their initial $25k-30k/box wasn't bad, it's just their CEO's pushy and arrogant sales guy style blocked the demo: the paperwork said I would have to pay full price even in case of a failed test if the box gets back to their office in CA after the 15th day (we are in NYC, just think about shipping times and cost!), not to mention sanctioning a mandatory purchase if tests results judged good... when I asked him to remove these ridiculous terms because I cannot commit to anything before I *KNOW* it is the right solution (I explained him that once I know I will sit down with my CFO and CEO and request the budget for the whole upgrade including his boxes) he immediately pulled out of the deal, citing (in an entertainingly patronizing tone) 'difficulties to justify the cost of supporting the evaluation', essentially admitting they do not have money to float a couple of demo unit around (and ours would've been an OLDER, small-sized unit.) szlevi fucked around with this message at 19:45 on Nov 27, 2012 |
# ¿ Nov 27, 2012 19:13 |
|
cheese-cube posted:Oh I see what they're saying, despite the stupid names they've used (Although that's probably just my completely irrational Dell hatred talking). I've worked with IBM SVC (SAN Volume Controller) before which which does the same sort of thing (They call it "external storage virtualisation"). This has nothing to do with Dell, every storage vendor I've talked to in the past 3-4 years called the old 'controller-plus-shelves-of-disk' setups "frame-based". Similarly every vendor called SAN systems that scale horizontally by simply multiplying the number of boxes (each with its own subsystem including controller/CPU/disks/whatever) "scale-out". And no, SVC is not scale-out at all, it's a storage virtualization product. In short it's just you, not Dell.
|
# ¿ Nov 27, 2012 20:00 |
|
Bitch Stewie posted:It's iSCSI not NFS but at a basic level a commodity Dell/HP running ESXi and a HP P4000 VSA doesn't sound a million miles off, though it only scales so far. Well, the VSA (a Proliant server, that is) could also run the hypervisor, giving us an almighty all-in-one building block... imagine deploying 10-15 of these cheap boxes, using a laptop to run the orchestration software (can be later also run in a VM)...
|
# ¿ Nov 27, 2012 21:23 |
|
NippleFloss posted:It still sounds like this is a heavily sequential workload which doesn't utilize the particular strengths of flash. And with frame sizes that large I'd imagine capacity would be a concern unless you're immediately vaulting most of this data onto slower storage after it's viewed once or twice. It's not my forte, but NetApp E-series is build for almost exactly this. It does very high ingest in a small form factor, and does it on spinning platters so you don't end up at huge $/TB multiple. Never seen anything economically feasible that would support this kind of bandwidth built on spinning disks (and I run an older 80-disks DDN S2A9550 here, topping at ~1.3GB/s, heh)... ...but Nimbus' building block is a 24-drive SSD-only unit, without the usual backend limitations (again, they designed & mfr their own drives, cutting out the controller, protocol etc fat) so we're talking about a LOT of bandwidth in a 2U box, giving me 5-10TB per box, for $30k and up. Say nI only get half of the bandwidth per box, around 5GB/s - that means 10GB/s will cost me the price of 2 boxes or, heck, calculate with 3 boxes... ...in comparison what do you think, how many spinning disk you need to sell me for that kind of bandwidth and at what price?
|
# ¿ Nov 28, 2012 00:02 |
|
Nomex posted:With Netapp you can have a large aggregate of spinning disk with some SSDs mixed in for cache. Blocks are moved up into the flash pool when they get busy, then get de-staged when they stop being accessed. It works kinda like their PAM cards, only it's read/write and aggregate specific rather than system wide. Yeah, that 'moving blocks'/tiering approach never worked, never will for this type of thing, I can tell you that already. As for being sequential - it's only per file but you have several users opening and closing plenty of different files etc so it's far from the usual video editing load. Size can vary from few gigabytes to 50-100GB per file (think of 4k and higher, RGBA fp32 etc) - we've developed our own raw format, wrote plugins for Fusion, Max etc so we're actually pretty flexible if it comes down to that... FWIW I have 2 Fuison-IO Duo cards, they were very fast when I bought them for $15k apiece, now they are just fast but the issue from day 1 is Windows CIFS: up to 2008 R2 (SMB2.1) CIFS is an utter piece of poo poo, it simply chops up everything into 1-3k size pieces so it pretty much destroys any chance of taking advantage of its bandwidth. Just upgraded my NAS boxes (Dell NX3000s) to Server 2012, I'll test SMB3.0 with direct FIO shares again - I'm sure it's got better but I doubt it's got that much better... Since going with an established big name would be very expensive (10GB/s!) as I see I have to choose between two approaches: 1. building my own Server 2012-based boxes eg plugging in 3-4 2GB/s or faster PCIe storage cards, most likely running the shebang as a file sharing cluster (2012 got a new active-active scale-out mode), hoping SMB3.0 got massively better 2. going with some new solution, coming from a new, small firm, hoping they will be around or bought up - and only, of course, after acquiring a demo unit to see real performance I can also wait until Dell etc buys up a company/rolls out something new but who knows when they will have affordable 10GB/s...? szlevi fucked around with this message at 21:52 on Nov 28, 2012 |
# ¿ Nov 28, 2012 21:50 |
|
Vanilla posted:I recall you mentioning isilon at some point in the past. Is it the cost that rules them out or a technical reason? Isilon never been cheap but now that it's carrying the EMC logo it would be even more expensive, not to mention the exodus of old hands from Isilon, the rash of issues in the the past year or so I heard of (from costumers and ex-eomployees) etc etc.
|
# ¿ Nov 28, 2012 21:55 |
|
NippleFloss posted:I'd be absolutely shocked if they can get their stated 12GB/s out of only 24 SSD drives. Assuming the data is protected at all you're losing at least a couple of those to parity drives, which means each SSD is doing more than 500 MB/s. You might get that out of a consumer quality SSD, but enterprise class SSDs only hit around 300 to 350 MB/s. There just aren't enough disks to get you to those throughput numbers with reliable hardware. It's possible that they are using consumer grade drives, but that would worry me, especially with a new vendor with no real track record. They make a lot of claims that border on magic, like 10 years of 1.2PB writes a weak without any performance loss, which just doesn't fit with the characteristics of SSD drives as they exist right now, especially not consumer grade drives. Could be true, but will they be around in 10 years to verify? I'd say they are probably using some sort of 8+2 raid sets, plus one or two hot spare... that gives you 2x10 drives so yes, 500MB/s - that's not unheard of and again, they claim that by having their own custom-designed (eg I know they have a small amount of memory built into every drive), custom manufactured SSDs they can gain enough extra bandwidth to get there (10Gb/s) with raid etc vs commercially available drives... who knows, could be true. quote:I know a E5460 box from NetApp can do about 3GB/s (this may be higher, theoretically, this is just the number I've seen when sizing for Lustre with a certain block size and stream count) in a 4u enclosure that includes redundant controllers and 60 7.2k 2/3TB NL-SAS drives. That'll get you around 80TB, give or take, with raid-6 and 2TB drives. I've got no idea on price though, since, as I said, I don't support these at all. Could be cheap of very expensive. It's probably less than the $8/GB raw that Nimbus gear lists at, but whether you need the extra capacity is another matter. I have to be able to feed 7-8 10gig-enabled people, that's ~8GB/s and I have no overhead left... with disks it'd be ridiculously more expensive, not to mention power, maintenance etc. We are also a Windows shop although I'm open for a proper (redundancy/HA, transparent Windows security support etc etc) non-Windows solution... Capacity does not matter, around 2-3TB is fine, 5-6TB would be downright future-proof, it's only for this purpose and I'd do my backend scripting/linking kungfu to hide it in our DFS hierarchy. quote:Anyway, my point wasn't that all flash arrays are bad, I'm just trying to understand who is using them and why. If your requirements are for a very high throughput low capacity solution in a small footprint then it might be the right move for you. SSD throughput is only about double spinning drive throughput, at best, but that might be enough difference to get to your magic number. I share your skepticism hence my request for a demo unit... quote:My guess is that Isilon would way too expensive and require too much gear to the to the 5 GB/s number he mentioned. I'd guess he'd be looking at 10 nodes, at minimum, to get to that number. Exactly.
|
# ¿ Nov 28, 2012 22:28 |
|
skipdogg posted:Paging Scott Allan Miller to this thread. He'll know just the thing! Oh please, no... not another pseudo-scientific/semi-pretentious explanation why everybody should build and run OpenFiler on el cheapo white boxes and how they are every bit of enterprise like brand-name units...
|
# ¿ Jun 5, 2013 16:29 |
|
Mierdaan posted:Bam. Ahahaha, awesome, saved.
|
# ¿ Jun 5, 2013 16:29 |
|
Amandyke posted:Pretty much this. Also helps to be near the end of the quarter or year. Triply helps if the sales guy is having trouble meeting his numbers and needs the sale... Plus if you trade in your equallogic you can get some credit for that. Potentially even more credit if the sales guy thinks he's getting a big win in pulling out other vendors gear. ^^^THIS - end of the year is THE BEST and end of a quarter is second best time to buy anything and all things you've mentioned will make them even sweeter. I literally almost never buy anything in-between quarters.
|
# ¿ Sep 17, 2013 17:43 |
|
Agrikk posted:My plan is to use Server 2012 R2 as my FC storage target, so it looks like these cards are supported on both sides of the link. All my tests resulted in utter crap performance when Windows was acting as a storage target. quote:If I get that working, I'll add in a silkworm switch to experiment with cluster storage groups and failover clustering under 2012. I run an active-active file sharing cluster on Storage Server 2012 Standard, feel free to ask if you want to know something.
|
# ¿ Sep 17, 2013 19:29 |
|
Quick question: for real-time file server share change auditing, which *reasonably* priced 3rd party tool (Windows Server 2012 is still utter junk when it comes to reading logs) should I be looking at...? Just two servers, half a dozen shares on each...
|
# ¿ Dec 10, 2013 02:05 |
|
skipdogg posted:In my experience the 80% margin number is usually on the software side of things. Hardware is probably around 60% depending. I personally shoot for at least 50% off MSRP, but it doesn't always happen. Commodity servers I can't seem to get below around 40, maybe 45% off MSRP if I spend 6 figures. I have an ongoing horror story about KACE but I promised to give them one more chance to right all the wrongs before I go nuclear online - which I will do, for sure, at least others would not buy into their BS anymore -; we will see, only few weeks left... szlevi fucked around with this message at 02:35 on Dec 10, 2013 |
# ¿ Dec 10, 2013 02:09 |
|
I have one more question: is anyone using ION Data Accelerator from Fusion-IO...? It seems to me like a competitive alternative to Violin's boxes (MUCH cheaper, even if you add the price of 2-3 FIO cards per server) - are those bandwidth numbers are for real..? Last time I've seen such claims - see http://www.fusionio.com/load/-media-/2griol/docsLibrary/FIO_DS_ION_DataAccelerator.pdf - was when I *almost* got a box from Nimbus Data, only to see the CEO himself (!) injecting some really nasty lending terms into the final doc he sent over for my signature, done in a very low-brow, disgustingly sneaky manner (and then even had the audacity of accusing me not having funds ready - while implicitly admitting he didn't even have available test boxes in circulation... clowns in the storage circus.)
|
# ¿ Dec 10, 2013 02:35 |
|
Dilbert As gently caress posted:What kind of workloads are you doing? We're a medical visualization firm, the actual workflow here would be high-end compositing using uncompressed data (we have our own raw file format), sometimes 4K, possibly even higher res in the future - to put it simple I want to reach at least 10-15 fps and frame size can go up as high as 128MB (those workstations have dual 10Gb adapters.) Currently all I'm doing is using two FIO cards as transprane t cahce, fronting two production volumes but 1. it's not nearly fast enough 2. it's limited to one card/volume 3. it's a clunky, manual process when you un-bind the crad from one volume and bind it to another one, depending on the location of the next urgent project folder...
|
# ¿ Dec 11, 2013 20:45 |
|
Dilbert As gently caress posted:The problem with doing something like FIO cards in the host is that data may/may not be actually written or modified back to your storage processors in time to provide a viable copy of media if a host fails. Could it potentially speed up things? Sure, just wait till a locks up, freezes, or the FIO card decides to take a dump. Doctors are going to be PISSED. We have artists and developers, we have no doctors. Our clients are big pharma and media firms so no end-users here. quote:before I got further are you using a VDI infrastructure like EPIC to do this on? No, it'd make no sense. We need 96-128GB RAM in these machines, they sport 6-core 3.33GHz Xeons etc (they were maxed-out Precision T7500s 3 years ago when I bought them.) quote:it's been a bit since I have worked with FIO cards so apologizes if I'm incorrect in their nature. I thought it goes without saying that I have the infrastructure in place... why would anyone install dual 10-gig NICs for gigabit switches? We run on 10Gb for 3-4 years now, few workstations. We can pull over 500MB/s from the FIO but that's enough for higher-res raw stuff. quote:What video cards do these workstations have? IGP may not cut it, but something like a 50 dollar AMD/Nvidia card in the remote workstation might. Something like an R7 240 with some ample video ram can drastically change these things. We have several plugins/tools we developed in CUDA so various NV cards: for the few we have high-end Quadros like K5000, rest are GTX desktop ones (480, 570, 770) - beyond CUDA compatibility the only thing that matters to us is the amount of memory, to work with larger datasets fast these tools load them into the video card's memory... I just bought a few GTX770 4GB for dirt cheap, they are great deals. quote:What is the latency of the client talking to the server hosting these images? Is the latency high on the Image server to datastore? These are simple project folders, on SMB3.0 file shares (Server 2012) and that's the issue. quote:What Datastores are you using? Flash accelerated storage works wonders for things, and flash storage is cheap. AGain, VDI is out of question - we need very high-end WS performance, that would make no sense.
|
# ¿ Dec 11, 2013 21:56 |
|
Dilbert As gently caress posted:Sorry I completely misread your "We're a medical visualization firm" as medical e.g. healthcare; It's been really an off day for me. I apologize. Oh, no, your questions were totally valid, I wasn't clear enough.
|
# ¿ Dec 12, 2013 19:29 |
|
The_Groove posted:That sounds fairly similar to a data analysis/visualization environment I used to manage, 12 clients with a ton of memory, 10GbE, and GPUs, so virtualization didn't make sense. We ran linux though, and ended up running IBM's GPFS filesystem with a 2MB block size on large DDN9550/DDN9900 storage systems (about 1.5PB total) with 8 NSD (I/O) servers in front of it, serving everything out over 10G. A single client could max out it's 10G card when doing sequential reads/writes and the 9900 alone could hit 4-5 GB/sec peaks under a heavy load. Granted GPFS is not even close to free and probably pretty expensive for a relatively "small" installation like that. It's more geared towards huge clusters and HPC, but drat did it rock for that environment. Nice, thanks for the info. Funny you said DDN - I have a 9550, it used to pump out ~1.5GB/s total, around 400-500MB/s per client but it's out of warranty for a while now and being a single-headed unit I don't dare to put it into production anymore.
|
# ¿ Dec 12, 2013 19:31 |
|
The problem with pnfs, lustre etc that 1. we are a small shop and hard to get proper support for something like that 2. we are a Windows shop and clients are totally not supported at best (or does not even exist)... We tested Stornext with our DDN back then and it sucked - I got better results with native NTFS, seriously (and that was a crap too.)
|
# ¿ Dec 12, 2013 19:38 |
|
|
# ¿ Apr 25, 2024 05:01 |
|
evol262 posted:Granted in some respects, and medical imaging is probably enormous files. You can find cheap arrays which'll max 10g with nearline SAS, clever caching, or SSDs. Doing it for multiple workstations simultaneously and with a relatively unknown I/O pattern is much harder. Yes and actually FIO just showcased their setup in the Summer, they had it running in their booth at SIGGRAPH: http://www.fusionio.com/siggraph-2013-demo/ This seems to be a LOT less complex for my team than a lustre setup (I'm handy with linux but pretty much that's it, nobody touches it unless necessary); essentially FIO applies some very clever caching all the way to the artist's WS, using block-level (IB) access... I probably wouldn't get those client cards but since I already have 2 FIO Duos that could drive the ioTurbine (newer version of my DirectCache) segment and I figured I might take a look how much those ioControl boxes (ex-Nexgen) cost, say, around 40-50TB... adding another pair of FIO cards in the other file server node could possibly bump me up to ~4GB/s, perhaps even further up...? Granted, it won't be as fast as the demo was and I will have to build it step-by-step but that's life.
|
# ¿ Dec 12, 2013 19:48 |