Re: [zfs-discuss] Ssd for zil on a dell 2950
Hello Greg, I'm curious how much performance benefit you gain from the ZIL accelerator. Have you measured that? If not, do you have a gut feel about how much it helped? Also, for what kind of applications does it help? (I know it helps with synchronous writes. I'm looking for real world answers like: "Our XYZ application was running like a dog and we added an SSD for ZIL and the response time improved by X%.") Of course, I would welcome a reply from anyone who has experience with this, not just Greg. Monish - Original Message - From: "Greg Mason" To: "HUGE | David Stahl" Cc: "zfs-discuss" Sent: Thursday, August 20, 2009 4:04 AM Subject: Re: [zfs-discuss] Ssd for zil on a dell 2950 Hi David, We are using them in our Sun X4540 filers. We are actually using 2 SSDs per pool, to improve throughput (since the logbias feature isn't in an official release of OpenSolaris yet). I kind of wish they made an 8G or 16G part, since the 32G capacity is kind of a waste. We had to go the NewEgg route though. We tried to buy some Sun-branded disks from Sun, but that's a different story. To summarize, we had to buy the NewEgg parts to ensure a project stayed on-schedule. Generally, we've been pretty pleased with them. Occasionally, we've had an SSD that wasn't behaving well. Looks like you can replace log devices now though... :) We use the 2.5" to 3.5" SATA adapter from IcyDock, in a Sun X4540 drive sled. If you can attach a standard sata disk to a Dell sled, this approach would most likely work for you as well. Only issue with using the third-party parts is that the involved support organizations for the software/hardware will make it very clear that such a configuration is quite unsupported. That said, we've had pretty good luck with them. -Greg -- Greg Mason System Administrator High Performance Computing Center Michigan State University HUGE | David Stahl wrote: We have a setup with ZFS/ESX/NFS and I am looking to move our zil to a solid state drive. So far I am looking into this one http://www.newegg.com/Product/Product.aspx?Item=N82E16820167013 Does anyone have any experience with this drive as a poorman’s logzilla? ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] ZFS, power failures, and UPSes
A related question: If you are on a UPS, is it OK to disable ZIL? The evil tuning guide says "The ZIL is an essential part of ZFS and should never be disabled." However, if you have a UPS, what can go wrong that really requires ZIL? Opinions? Monish - Original Message - From: "Ross" To: Sent: Tuesday, June 30, 2009 3:04 PM Subject: Re: [zfs-discuss] ZFS, power failures, and UPSes I've seen enough people suffer from corrupted pools that a UPS is definitely good advice. However, I'm running a (very low usage) ZFS server at home and it's suffered through at least half a dozen power outages without any problems at all. I do plan to buy a UPS as soon as I can, but it seems to be surviving very well so far. -- This message posted from opensolaris.org ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] compression at zfs filesystem creation
Unless you're in GIMP working on JPEGs, or doing some kind of MPEG video editing--or ripping audio (MP3 / AAC / FLAC) stuff. All of which are probably some of the largest files in most people's homedirs nowadays. indeed. I think only programmers will see any substantial benefit from compression, since both the code itself and the object files generated are easily compressible. If we are talking about data on people's desktops and laptops, yes, it is not very common to see a lot of compressible data. There will be some other examples, such as desktops being used for engineering drawings. The CAD files do tend to be compressible and they tend to be big. In any case, the really interesting case for compression is for business data (databases, e-mail servers, etc.) which tends to be quite compressible. ... I'd be interested to see benchmarks on MySQL/PostgreSQL performance with compression enabled. my *guess* would be it isn't beneficial since they usually do small reads and writes, and there is little gain in reading 4 KiB instead of 8 KiB. OK, now you have switched from compressibility of data to performance advantage. As I said above, this kind of data usually compresses pretty well. I agree that for random reads, there wouldn't be any gain from compression. For random writes, in a copy-on-write file system, there might be gains, because the blocks may be arranged in sequential fashion anyway. We are in the process of doing some performance tests to prove or disprove this. Now, if you are using SSDs for this type of workload, I'm pretty sure that compression will help writes. The reason is that the flash translation layer in the SSD has to re-arrange the data and write it page by page. If there is less data to write, there will be fewer program operations. Given that write IOPS rating in an SSD is often much less than read IOPS, using compression to improve that will surely be of great value. At this point, this is educated guesswork. I'm going to see if I can get my hands on an SSD to prove this. Monish what other uses cases can benefit from compression? -- Kjetil T. Homme Redpill Linpro AS - Changing the game ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] compression at zfs filesystem creation
Hello Richard, Monish Shah wrote: What about when the compression is performed in dedicated hardware? Shouldn't compression be on by default in that case? How do I put in an RFE for that? Is there a bugs.intel.com? :-) I may have misled you. I'm not asking for Intel to add hardware compression. Actually, we already have gzip compression boards that we have integrated into OpenSolaris / ZFS and they are also supported under NexentaStor. What I'm saying is that if such a card is installed, compression should be enabled by default. NB, Solaris already does this for encryption, which is often a more computationally intensive operation. Actually, compression is more compute intensive than symmetric encryption (such as AES). Public key encryption, on the other hand, is horrendously compute intensive, much more than compression or symmectric encryption. But, nobody uses public key encryption for bulk data encryption, so that doesn't apply. Your mileage may vary. You can always come up with compression algorithms that don't do a very good job of compressing, but which are light on CPU utilization. Monish I think the general cases are performed well by current hardware, and it is already multithreaded. The bigger issue is, as Bob notes, resource management. There is opportunity for people to work here, especially since the community has access to large amounts of varied hardware. Should we spin up a special interest group of some sort? -- richard ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] compression at zfs filesystem creation
Hello, I would like to add one more point to this. Everyone seems to agree that compression is useful for reducing load on the disks and the disagreement is about the impact on CPU utilization, right? What about when the compression is performed in dedicated hardware? Shouldn't compression be on by default in that case? How do I put in an RFE for that? Monish On Mon, 15 Jun 2009, dick hoogendijk wrote: IF at all, it certainly should not be the DEFAULT. Compression is a choice, nothing more. I respectfully disagree somewhat. Yes, compression shuould be a choice, but I think the default should be for it to be enabled. I agree that "Compression is a choice" and would add : Compression is a choice and it is the default. Just my feelings on the issue. Dennis Clarke ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] Asymmetric mirroring
Hello, Thanks to everyone who replied. Dan, your suggestions (quoted below) are excellent and yes, I do want to make this work with SSDs, as well. However, I didn't tell you one thing. I want to compress the data on the drive. This would be particularly important if an SSD is used, as the cost per GB is high. This is why I wanted to put it in a zpool. Before somebody points out that compression with increase the CPU utilization, I'd like to mention that we have hardware accelerated gzip compression technology already working with ZFS, so the CPU will not be loaded. I'm also hoping that write IOPS will improve with compression, because more writes can be combined into a single block of storage. I don't know enough about ZFS allocation policies to be sure, but we'll try to run some tests. It looks like, for now, the mirror disks will also have to be SSDs. (Perhaps raidz1 will be OK, instead.) Eventually, we will look into modifying ZFS to support the kind of asymmetric mirroring I mentioned in the original post. The other alternative is to modify ZFS to compress L2ARC, but that sounds much more complicated to me. Any insights from ZFS developers would be appreciated. Monish Monish Shah CEO, Indra Networks, Inc. www.indranetworks.com Use the SAS drives as l2arc for a pool on sata disks. If your l2arc is the full size of your pool, you won't see reads from the pool (once the cache is primed). If you're purchasing all the gear from new, consider whether SSD in this mode would be better than 15k sas. -- This message posted from opensolaris.org ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
[zfs-discuss] Asymmetric mirroring
Hello everyone, I'm wondering if the following makes sense: To configure a system for high IOPS, I want to have a zpool of 15K RPM SAS drives. For high IOPS, I believe it is best to let ZFS stripe them, instead of doing a raidz1 across them. Therefore, I would like to mirror the drives for reliability. Now, I'm wondering if I can get away with using a large capacity 7200 RPM SATA drive as mirror for multiple SAS drives. For example, say I had 3 SAS drives of 150 GB each. Could I take a 500 GB SATA drive, partition it into 3 vdevs and use each one as a mirror for one SAS drive? I believe this is possible. The problem is in performance. What I want is for all reads to go to the SAS drives so that the SATA drive will only see writes. I'm hoping that due to the copy-on-write nature of ZFS, the writes will get bunched into sequential blocks, so write bandwidth will be good, even on a SATA drive. But, the reads must be kept off the SATA drive. Is there any way I can get ZFS to do that? Thanks, Monish ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] Can this be done?
Hello David and Michael, Well I might back up the more important stuff offsite. But in theory it's all replaceable. Just would be a pain. And what is the cost of the time to replace it versus the price of a hard disk? Time ~ money. This is true, but there is one counterpoint. If you do raidz2, you are definitely paying for extra disk(s). If you stay with raidz1, the cost of the time to recover the data would be incurred if and only if he has a failure in raidz1 followed by a second failure during the re-build process. So, the statistically correct thing to do is to multiply the cost of recovery by the probability and see if that exceeds the cost of the new drives. To be really accurate, the cost of raidz2 option should also include the cost of moving the data from the existing raidz1 to the new raidz2 and then re-formatting the raidz1 into raidz2. However, all this calculating is probably not worthwhile. My feeling is: it's just a home video server and Michael still has the original media (I think). Raidz1 is good enough. Monish ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Re: [zfs-discuss] Encryption through compression?
Hello Darren, Monish Shah wrote: Hello everyone, My understanding is that the ZFS crypto framework will not release until 2010. That is incorrect information, where did you get that from ? It was in Mike Shapiro's presentation at the Open Solaris Storage Summit that took place a couple of weeks ago. Perhaps I mis-read the slide, but I'm pretty sure it listed encryption as a feature for 2010. ... 3. There is no key management framework. That is impossible there has to be key management somewhere. What I meant was, the compression framework does not have key management framework. Using our hardware (which I mentioned later in my mail), the key management would come with the hardware, since we store keys in the hardware. We provide a utility to manage the keys stored in the hardware. ... If it is specific to your companies hardware I doubt it would ever get integrated into OpenSolaris particularly given the existing zfs-crypto project has no hardware dependencies at all. The better way to use your encryption hardware is to get it plugged into the OpenSolaris cryptographic framework (see the crypto project on OpenSolaris.org) That was precisely what I want thinking originally. However, if it is out in 2010, there is temptation to do our own project, which I thought could be done in a couple of months. (In light of your comment below, my estimate may have been wildly optimistic, but the foregoing is merely an explanation of what I was thinking.) Does anyone see any problems with this? There are probably various gotchas here that I haven't thought of. If you can think of any, please let me know. The various gotchas are the things that have been taking me and the rest of the ZFS team a large part of the zfs-crypto project to resolve. It really isn't as simple as you think it is - if it were then the zfs-crypto project would be done by now! If you really want to help get encryption for ZFS then please come and join the already existing project rather than starting another one from scratch. If the schedule is much sooner than 2010, I would definitely do so. What is your present schedule estimate? -- Darren J Moffat Monish ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
[zfs-discuss] Encryption through compression?
Hello everyone, My understanding is that the ZFS crypto framework will not release until 2010. In light of that, I'm wondering if the following approach to encryption could make sense for some subset of users: The idea is to use the compression framework to do both compression and encryption in one pass. This would be done by defining a new compression type, which might be called "compress-encrypt" or something like that. There could be two levels, one that does both compress and encrypt and another that does encrypt only. I see the following issues with this approach: 1. ZFS compression framework presently takes compressed data only if there was at least 12.5% reduction. For data that didn't compress, you would wind up storing it unencrypted, even if encryption was on. 2. Meta-data would not be encrypted. I.e., even if you don't have the key, you will be able to do directory listings and see file names, etc. 3. There is no key management framework. I would deal with these as follows: Issue #1 can be solved by changing ZFS code such that it always accepts the "compressed" data. I guess this is an easy change. Issue #2 may be a limitation to some and feature to others. May be OK. Issue #3 can be solved using encryption hardware (which my company happens to make). The keys are stored in hardware and can be used directly from that. Of course, this means that the solution will be specific to our hardware, but that's fine by me. The idea is that we would do this project on our own and supply this modified ZFS with our compression/encryption hardware to our customers. We may submit the patch for inclusion in some future version of OS, if the developers are amenable to that. Does anyone see any problems with this? There are probably various gotchas here that I haven't thought of. If you can think of any, please let me know. Thanks, Monish Monish Shah CEO, Indra Networks, Inc. www.indranetworks.com ___ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss