|
You will have 16TB of raw space (8 disks at 2TB each) but that is going to drop to 10-11GB once it's formatted and you subtract a single spare. If you're going active/active, that will be like 5TB per controller, or you can pool it all together for one 10-11GB aggregate on your primary controller. NippleFloss says a/p is less than ideal, but I actually prefer it for the smaller deployments. By putting more disks in a single aggregate, you get better performance since you're striping across more disks, and you're definitely not going to be capping out a single 2240's capabilities with 12x SATA disks. You get more flexibility since you have all of the storage available to you in one pool. Makes configuring way easier too, since you essentially just act as if you have 1 controller and point everything at it, rather than trying to remember which services are hosted by which storage. re: spares, since you have a NetApp support agreement anyway, I like just keeping a single spare for both heads. You can assign it to whichever head has the first failure, and your new disk will be there shortly to fill in.
|
#
?
Mar 28, 2012 22:39
|
|
|
|
| # ? May 10, 2024 19:21 |
|
|
madsushi posted:You will have 16TB of raw space (8 disks at 2TB each) but that is going to drop to 10-11GB once it's formatted and you subtract a single spare. If you're going active/active, that will be like 5TB per controller, or you can pool it all together for one 10-11GB aggregate on your primary controller. I don't have an issue with unbalanced configurations. I agree that it's often a lot more useful to give most of the disk to just one controller when you only have 12 disks. The problem is that it's still an active/active pair in that scenario. You still have to provide at minimum two disks to the "passive" controller to create an aggregate and those disks aren't contributing anything to your usable space if you treat that node as passive by just ignoring it. On the subject of spares it's nice to have at least one on each controller so DOT always has a spare available to do disk pre-fails which saves a rebuild and lessens the time you're running degraded. But, as you said, in a disk limited scenario that isn't always viable.
|
|
#
?
Mar 28, 2012 23:30
|
|
|
NippleFloss posted:I don't have an issue with unbalanced configurations. I agree that it's often a lot more useful to give most of the disk to just one controller when you only have 12 disks. The problem is that it's still an active/active pair in that scenario. You still have to provide at minimum two disks to the "passive" controller to create an aggregate and those disks aren't contributing anything to your usable space if you treat that node as passive by just ignoring it. Actually it's a minimum of 3 disks to the passive controller, which is even worse. With RAID-DP, disk 1 and disk 2 are parity/dual-parity, and you need disk 3 to actually store the data.
|
|
#
?
Mar 29, 2012 01:15
|
|
|
madsushi posted:Actually it's a minimum of 3 disks to the passive controller, which is even worse. With RAID-DP, disk 1 and disk 2 are parity/dual-parity, and you need disk 3 to actually store the data. A lot of people who do active passive do RAID 4 rather than DP on the "passive" node. DP makes little sense when you only have a single data disk and it only holds configuration information for a mode with no user data.
|
|
#
?
Mar 29, 2012 01:26
|
|
|
I sent NetApp a follow up email asking them about the real amount of available storage, making sure that we're not missing any software that we needed, and also mentioning people getting 24X2TB configurations for around the same price we were quoted for 12X2TB. I also followed up with Nimble and they'll have a quote for me tomorrow. Hopefully, I can either drive the price down or the storage up. I think 12TBs should work for us, but I know people tend to underestimate their storage needs, so if I can get more storage for the same price, then I'll find a use for it.
|
|
#
?
Mar 29, 2012 01:31
|
|
|
NippleFloss posted:A lot of people who do active passive do RAID 4 rather than DP on the "passive" node. DP makes little sense when you only have a single data disk and it only holds configuration information for a mode with no user data. Doesn't RAID 4 require 3 disks? Or am I missing something?
|
|
#
?
Mar 29, 2012 01:34
|
|
|
Aniki, you are experiencing a bad partner, sorry  Everything NippleFloss has said is right on the money, and you should absolutely push to get the complete bundle. I would also say that the price you've been quoted is quite high.
|
|
#
?
Mar 29, 2012 03:01
|
|
|
madsushi posted:Doesn't RAID 4 require 3 disks? Or am I missing something? Raid 4 is single parity with a dedicated parity disk. A 2 disk raid aggregate is 1 data disk and 1 parity disk.
|
|
#
?
Mar 29, 2012 04:22
|
|
|
I wish I could find it but there was an IBM redbook which broke down the additional penalies with ONTAP, a few more % in additonal storage, with flexvolumes for instance. By the way if you didn't know it, IBM's N series are really rebadged netapp filers. And there are a lot of useful documents. http://publib-b.boulder.ibm.com/Red...ategory=Storage
|
|
#
?
Mar 29, 2012 07:06
|
|
|
marketingman posted:Aniki, you are experiencing a bad partner, sorry I was expecting the initial quote to be high, so I'll see how they respond. Do you have any idea on what price range I should expect to see? You can PM me if you'd rather not share that information in the thread. I'll definitely push to bring the price down and get the complete software bundle, but it would be good to have an idea of what a SAN like that should actually cost.
|
|
#
?
Mar 29, 2012 11:48
|
|
|
NippleFloss posted:I hate to say it, but their usable space calculations are misleading. You start with 12 disks. In an active/active configuration you must have at minimum two raid groups, as each controller requires it's own aggregate. With raid-DP each raid group uses two parity disks, so that knocks your 12 disks down to 8. Best practices are to leave at least one hot spare available on each controller, which would drop you down to six data disks. I'm glad someone else is breaking this down: I knew 16TB of usable allocatable space at that pricepoint didn't sound right at all.
|
|
#
?
Mar 29, 2012 14:04
|
|
|
Aniki posted:For people more familiar with pricing of SAN storage is $34k a good price for a NetApp 2240, w/ 12 2TB HDs, 2 controller cards, and 10GB Ethernet? This is the end of the NetApp fiscal year, beat them up on the price and I'm sure you can get it lower.
|
|
#
?
Mar 29, 2012 15:12
|
|
|
j3rkstore posted:This is the end of the NetApp fiscal year, beat them up on the price and I'm sure you can get it lower. I'm trying a couple things with them, yesterday I told them to either give me a lower price ont he 12X2TB configuration or put together a quote for the 24X2TB configuration that is in line with what other people in this thread have paid. They're currently working on a quote for the 24X2TB configuration, which they claim has deeper discounts associated with it. I'm also pushing them on the software and it sounds like it's one of those things where if you push them on it, then they'll just include it rather than risk losing the sale. I've also heard that it's cheaper to buy the QLogic 10GBe card/controller than to get the card from NetApp, since it is the same card as what NetApp provides, but NetApp just rebrands it and adds markup. I'll see what happens, but I'm also concurrently getting quotes from Nimble and I can point out to NetApp that Nimble provides all of their software with their equipment. I've been looking into arranging a trial with Nimble, but basically their trial agreement is a contingent sale based on certain goals being met within a 10-day timespan. Does Nimble experience similar losses in storage capacity to NetApp? Is 8TB useable actually 8TB or would it only give me 6 or 7TB in real space. I know they have compression, but we're not expecting significant gains for most of our data (archived calls).
|
|
#
?
Mar 29, 2012 18:23
|
|
|
Aniki posted:I'm trying a couple things with them, yesterday I told them to either give me a lower price ont he 12X2TB configuration or put together a quote for the 24X2TB configuration that is in line with what other people in this thread have paid. They're currently working on a quote for the 24X2TB configuration, which they claim has deeper discounts associated with it. I'm also pushing them on the software and it sounds like it's one of those things where if you push them on it, then they'll just include it rather than risk losing the sale. Nimble likely includes all software, but they lack some features that NetApp has so its not apples to apples. For instance I don't believe they have the ability to perform automated restore of single files from within a VMDK. If you wanted to restore only certain files, rather than the whole vmdk, you would need to clone the LUN from the snapshot, export it to VMware, mount it as a new datastore, attach the vmdk from snapshot to the running host, copy the data off, and then reverse the process for cleanup. That isn't too bad every once in a while, but it's not ideal and there is no option for some level of user self service. The NetApp tool automates much of that workflow and provides a limited self-service option that requires less admin intervention. It's also all done through vSphere. I'm not sure where Nimble stands on application backup, but NetApp can do online, application consistent backup of SQL, Oracle, Exchange, Sharepoint and a few others. I don't believe Nimble currently does any of those.
|
|
#
?
Mar 29, 2012 21:04
|
|
|
Aniki posted:netapp quote shenanigans. I've tried to get quotes for Netapp from two VARs in my state. The first partner I tried kept quoting things way out of our price range and being far too aggressive, leaving huge amounts of voicemails/emails that just put us off dealing with them. The other partner calls or emails me every now and then talking about the quote we're going to get but not actually providing it. In one email (sent after the quote was apparently ready) they wanted to know what type of power cables we might need, another was essentially "come drive two and a half hours to attend a sales pitch about other tech we might sell you, oh, and we might show you the Netapp quote also." Are all Netapp resellers this weird because the two I've dealt with so far have been shady as gently caress. Did you have any luck with Equallogic?
|
|
#
?
Mar 29, 2012 23:05
|
|
|
ozmunkeh posted:I've tried to get quotes for Netapp from two VARs in my state. The first partner I tried kept quoting things way out of our price range and being far too aggressive, leaving huge amounts of voicemails/emails that just put us off dealing with them. Haven't tried Equallogic yet, I've been focusing on NetApp and Nimble, but after I get updated quotes tomorrow, then I may look into them. It looks like we may not need to rush into virtualization with the new phone system, but I've seen enough benefits of SANs that it is pretty clear we need one regardless of whether we choose to virtualize now or down the road. My NetApp reps have at least been giving me quotes, they've just been highballing me and not including all of the software that they should. Aniki fucked around with this message at 00:09 on Mar 30, 2012 |
|
#
?
Mar 30, 2012 00:05
|
|
|
If you don't have a dedicated storage admin I really can't recommend Equallogic enough. We went from them to an EMC VNX 5300 about 3 months ago and I am miserable. Everything is much more complicated on the EMC box and there isn't THAT big of a difference in features to make up for it. Maybe if my day to day job was dealing with SANs it would be different, but as it is now it takes up way too much of my time. Can't imagine NetApp is any different.
|
|
#
?
Mar 30, 2012 00:07
|
|
|
Internet Explorer posted:If you don't have a dedicated storage admin I really can't recommend Equallogic enough. We went from them to an EMC VNX 5300 about 3 months ago and I am miserable. Everything is much more complicated on the EMC box and there isn't THAT big of a difference in features to make up for it. Maybe if my day to day job was dealing with SANs it would be different, but as it is now it takes up way too much of my time. Can't imagine NetApp is any different. Most modern SAN devices, especially if they don't do FC, are pretty easy to manage. EMC just sucks.
|
|
#
?
Mar 30, 2012 01:39
|
|
|
Internet Explorer posted:If you don't have a dedicated storage admin I really can't recommend Equallogic enough. We went from them to an EMC VNX 5300 about 3 months ago and I am miserable. Everything is much more complicated on the EMC box and there isn't THAT big of a difference in features to make up for it. Maybe if my day to day job was dealing with SANs it would be different, but as it is now it takes up way too much of my time. Can't imagine NetApp is any different.
|
|
#
?
Mar 30, 2012 07:40
|
|
|
Well poo poo, the same thing that happened to one of our EMC SANs in Los Angeles has happened in our Oakland office now. It hit a bad sector and took down the array. I was up until 2am working with support, but they don't want to do anything because apparently when it hit a bad sector, it also caused the management agents on the storage processors to crash, so they can't see what's going on with the backend. This is insanity, and I'm very glad I had already moved half the stuff to our new Compellent SAN. Unfortunately not the really important stuff yet, but I'm definitely through with EMC forever.
|
|
#
?
Mar 30, 2012 16:54
|
|
|
Well, that's terrifying. What model and OS version? You know, when you're done picking up the pieces of an exploded SAN.
|
|
#
?
Mar 30, 2012 17:54
|
|
|
Serfer posted:It hit a bad sector and took down the array.
|
|
#
?
Mar 30, 2012 18:14
|
|
|
Internet Explorer posted:Well, that's terrifying. What model and OS version? You know, when you're done picking up the pieces of an exploded SAN. evil_bunnY posted:There are no words, basically. What model? It's an NX4 (AX4 with an NFS frontend) version 5.4.something, yes lowend stuff, but it was good enough at the time. Now that we need more, we've ordered Compellent. Last time this happened they basically said, "lol this happens, buy a VNX to fix the problem!" See these previous posts: http://forums.somethingawful.com/showthread.php?goto=post&postid=397522968 http://forums.somethingawful.com/showthread.php?goto=post&postid=398198862
|
|
#
?
Mar 30, 2012 18:45
|
|
|
So I'm a little lost trying to find what I need, so maybe someone can point me in the right direction. We need to replace our NAS for our backup server. What we're looking for is: ~8 TB would work, room to grow would be nice. iSCSI Raid 6 would be nice, 5 would work. Other than that, it can be dumb as a brick - No replication, dedupe, etc. Price is kinda an issue, and we want to aim at around $5,000. Is that even possible? Anything else in need of consideration?
|
|
#
?
Mar 30, 2012 22:37
|
|
|
I really like our Synology NAS. There should be one in your price that can handle that. Not 100% sure with the markup on disks right now, but shouldn't be impossible.
|
|
#
?
Mar 30, 2012 22:50
|
|
|
does $180k sound normal for a netapp 3240 HA pair, 2 shelves of 1tb sata disks, complete pack, and 3 years of maintenance? I am assuming that I cannot replace a 3140 ha pair with a 3240 ha pair without any downtime, but I can just replace the heads and power up, right?
|
|
#
?
Mar 30, 2012 23:18
|
|
|
Oh, has anyone lately mentioned that Pillar is overpriced poo poo? OK, Pillar is overpriced poo poo. Might as well be in the title. I hated it 3 years ago and our customers have pretty much all dumped them as well. Hilarious. Nice try, Ellison.
|
|
#
?
Mar 30, 2012 23:38
|
|
|
adorai posted:does $180k sound normal for a netapp 3240 HA pair, 2 shelves of 1tb sata disks, complete pack, and 3 years of maintenance? Yes, the upgrade could just be done as a head swap, though there is some prep work involved. You need to upgrade the 3140 cluster to the same version the 3240 shipped with and if you use FC you would need to update zoning to reflect new WWNs. You also have to change the disk assignment in maintenance mode and you'll need to set the appropriate root vol. It's not difficult and it's pretty quick, but it's not wholly trivial.
|
|
#
?
Mar 30, 2012 23:57
|
|
|
NippleFloss posted:It's not difficult and it's pretty quick, but it's not wholly trivial.
|
|
#
?
Mar 31, 2012 00:09
|
|
|
adorai posted:For someone who would do this literally once every 3 years, is it worth paying someone to come onsite and not gently caress it up? (We don't use FC, so at least we have that going for us) I would recommend it. The chance of data loss is pretty low unless you do something incredibly silly, but you could definitely end up causing an extended outage and having to call in help anyway. Bundling professional services to perform the upgrade should be included as part of the purchase. It shouldn't add that much cost as the actual work should only take a few hours. YOLOsubmarine fucked around with this message at 00:46 on Mar 31, 2012 |
|
#
?
Mar 31, 2012 00:43
|
|
|
NippleFloss posted:Bundling professional services to perform the upgrade should be included as part of the purchase. It shouldn't add that much cost as the actual work should only take a few hours.
|
|
#
?
Mar 31, 2012 01:52
|
|
|
Orcs and Ostriches posted:So I'm a little lost trying to find what I need, so maybe someone can point me in the right direction. I've had good luck with the Qnap TS-859u RP+ (or the TS-459u RP+). You can get a TS-859u RP+ with 8 3TB drives for under $5k. Supports iSCSI & NFS.
|
|
#
?
Mar 31, 2012 02:13
|
|
|
adorai posted:I figured it would probably be $2000 plus travel, lodging, and meals since we'll want to do it after hours for obvious reasons. Seem reasonable? Negotiate it down to nothing, your spending $200k if they want the business they will eat the cost.
|
|
#
?
Mar 31, 2012 04:31
|
|
|
So, where would one go to get rid of existing SAN gear? We've got some Isilon units that are being replacing by another vendor's solution this year, and they will be end-of-life next year, so it doesn't seem to make sense to continue using it in production. But on the other hand, I've got no idea what to do with it.
|
|
#
?
Apr 2, 2012 17:05
|
|
|
Fourteen posted:So, where would one go to get rid of existing SAN gear? We've got some Isilon units that are being replacing by another vendor's solution this year, and they will be end-of-life next year, so it doesn't seem to make sense to continue using it in production. But on the other hand, I've got no idea what to do with it.
|
|
#
?
Apr 2, 2012 17:14
|
|
|
Another amusing EMC anecdote. Standard disclaimer applies, I am not a SAN admin, I am new to EMC gear, what I say could be incorrect, etc. etc. On our VNX 5300 we use CIF shares off the SAN. We have Checkpoints (snapshots) in place. When we first set it up we did a daily snapshot and configured it to keep 5 snapshots. You can do all this through the GUI when you set it up. If you go in to edit that schedule, instead of create a new one, you get the same screen to set it all up, minus the part asking you how many snapshots you want to keep. If you want to edit that number, you have to log in through the command line. Okay, no big deal. I'll log in through the command line. There is no command to edit the number of snapshots you want to keep. You have to run a command to dump the info about the schedule, copy that info, run a command to delete the schedule, then run a command to create the schedule while tweaking the number of snapshots you want to keep. Really? In 2012? From the biggest player in SAN / Virtualization? :emc: [Edit: Oh, I'm sorry. There is a command to modify the Checkpoint schedule in the command line. But you still cannot modify the number of snapshots kept with it.] Internet Explorer fucked around with this message at 19:08 on Apr 2, 2012 |
|
#
?
Apr 2, 2012 18:52
|
|
|
Serfer posted:It's an NX4 (AX4 with an NFS frontend) version 5.4.something, yes lowend stuff, but it was good enough at the time. Now that we need more, we've ordered Compellent. So after spending 14 hours Saturday and 10 Sunday in the Oakland office, they say "Oh it's a software problem, you need an upgrade" "Ok, great, except I requested this upgrade twice before, said I'm perfectly willing to do it myself, but you won't give me access to the files needed to upgrade, and you guys totally ignored my two previous requests." "But this time we have managers in on the issue" All you need to have happen if you want something done is to have your SAN totally stop working. The machine is still throwing unrecoverable parity errors (they say they're false, I doubt it), and storage processor B is crashing every couple hours, but hey, it's accessible, problem solved.
|
|
#
?
Apr 2, 2012 20:58
|
|
|
Serfer posted:So after spending 14 hours Saturday and 10 Sunday in the Oakland office, they say "Oh it's a software problem, you need an upgrade"
|
|
#
?
Apr 3, 2012 04:43
|
|
|
I don't know if this is the best place do ask - please do direct me if there's somewhere better. We're running servers using SuperMicro X8DTU-6TF+ boards which have onboard LSI 9260-8i. We keep seeing these errors: [164188.395505] sd 4:2:5:0: [sdg] megasas: RESET -148955001 cmd=0 retries=0 [164188.395509] megasas: [ 0]waiting for 1 commands to complete [164189.403736] megasas: reset successful [164199.387535] sd 4:2:5:0: [sdg] megasas: RESET -148955009 cmd=0 retries=0 [164199.387540] megasas: [ 0]waiting for 1 commands to complete [164200.396456] megasas: reset successful [164210.380050] sd 4:2:5:0: [sdg] megasas: RESET -148955010 cmd=0 retries=0 [164210.380053] megasas: [ 0]waiting for 1 commands to complete [164211.387734] megasas: reset successful [164221.373243] sd 4:2:5:0: [sdg] megasas: RESET -148955013 cmd=0 retries=0 [164221.373247] megasas: [ 0]waiting for 1 commands to complete [164222.380128] megasas: reset successful Does anyone know what might be causing this?
|
|
#
?
Apr 3, 2012 07:59
|
|
|
|
| # ? May 10, 2024 19:21 |
|
|
It's the megasas driver being too verbose. I don't think those are actually errors.
|
|
#
?
Apr 3, 2012 19:48
|
|