My new solid state drive raid 0 array – sucks!

Published on January 11th, 2009

Posted by Marc Perkel in General, Research, Technology

http://www.dvorak.org/blog/images/raid0.jpg

Solid State Drives (SSD) are getting cheap and fast. So I’m putting 4 of them together in a Raid 0 array for a super fast MySQL server. The drives are just 30 gigs but cost only $75 each at NewEgg. Trying a 4k chunk size as these drives are unbuffered. I think that’s the sweet spot.

REVISED

Bottom line – it doesn’t work. I’m still investigating why but instead of being fast it’s extremely slow. Just formatting it with mke2fs takes 4 minutes and 54 seconds. These drives are made by OCZ and I think they are cooking the numbers on the flash drive speed. I think that the real numbers are more than 10 times slower than they specify.

49 users responded in " My new solid state drive raid 0 array – sucks! "

Subscribe to this post comment rss or trackback url

# 1 amodedoma said, on January 8th, 2009 at 7:04 pm

Cool project! MORE DETAILS! If not here post it on hack-a-day, please. Oh and please include Benchmarks! Been waiting for prices to come down to do something similar…

# 2 Rob Alter said, on January 8th, 2009 at 7:22 pm

Remember that SSD drives have limited write cycles. Flash-memory cells will often wear out after 1,000 to 10,000 write cycles for MLC, and up to 100,000 write cycles for SLC. Many log files, file allocation tables, and other commonly used parts of the file system exceed this quite quickly.

# 3 Somebody_Else said, on January 8th, 2009 at 7:33 pm

Nice. I just picked up a G. Skill 64 GB drive for ~$130 from Newegg.

The documentation claims that it has a MTFB of 1,500,000 hours and can endure 50 GB of writes/erases per day for 35.07 years . We’ll see.

I’m happy with the performance. Vista is nice and snappy and applications open very quickly. It does slow down significantly while writing, but its read speeds are fantastic.

# 4 James Hill said, on January 8th, 2009 at 7:36 pm

Very cool. Can you post some performance measurements?

# 5 Rob Alter said, on January 8th, 2009 at 8:23 pm

Somebody_Else:

I think MTBF (Mean Time Between Failure) only refers to failure, not write cycles.

All this marketing propaganda gets confusing…

# 6 Engineer said, on January 8th, 2009 at 8:32 pm

Unfortunately, all this data needs to be looked at with a lot of care. The MTBF does not have any meaning at all, it could as well last 10 minutes.
Also consider that those drives use MLC and not SLC chips. The random write is much worse than the stated 80 Mbps (which means Mbit per second and is wrong in this case. It should be MB/s = Megabyte per second). Besides, the prices are dropping, along with the quality.

# 7 Somebody_Else said, on January 8th, 2009 at 8:48 pm

#4 and #5
The documentation specifically lists the “drive endurance” as 50 GB of write/erase cycles per day for 35.07 years. The MTBF is a separate section.

Obviously nobody knows how long these drives will actually last, but the wear leveling they’re using on these first generation drives is supposed to be very good at extending the life of the flash cells.

I don’t think these drives are quite ready for widespread adoption because of the slow write speeds, but I could see them starting to catch on this year as prices decline and performance improves.

# 8 Mark Derail said, on January 8th, 2009 at 8:51 pm

Ugly couch, cute drives.
Flash disks have limited lifespans, SSD’s don’t. Any SSD is more reliable than a regular SATA HD.

Your chunk size should reflect the largest row size of the database table that has the most daily delta, times 2.

So 4k is most likely still too big for optimum performance, but a good overall choice.

FWIW, I have a table that grows 10k rows per day, each row is 496 bytes. Performance with chunks at 512 bytes or 1024 bytes were identical. Some gain at 2048, due to read-ahead, no gain at 4k versus 2k.
At 8k it was just like 4k, I gave up the database benchmark.

Was timing doing 1 million inserts, then 10,000 selects, worst case scenario.

I was doing SCSI Raid-5 across six HD’s. Performance with six 7200 rpm versus 10k versus 15k, the 15k’s were not twice as fast, only slightly better than the 10k’s.

Wasn’t worth twice the price for 15k vs 10k.

# 9 syrinx said, on January 8th, 2009 at 9:35 pm

@Marc Perkel

4k chunk “size.”

# 10 Someone said, on January 8th, 2009 at 10:04 pm

These may have been the best comments I’ve ever read here.

# 11 JAJansenJr said, on January 9th, 2009 at 12:28 am

I once wrote a database application for (remember this one?) a 486 PC. It barely crawled and was unusable. I rewrote it in VB6 using a technology called DAO (data access objects) that came with VB6.
At the beginning of the day I would scan the hard drive using DAO to look for physical record location
and product number information, and built a RAM table with this information. During the day I would search in the RAM table, pass the location info to a DAO routine, and pluck the record off the hard drive. What was an unusable application became an application that performed “instantly”. The point of this story is that excellent performance can be obtained if you move slow operations (such as searching on a hard drive) into RAM and just use the hard drive as a “clothes closet” (ever try to search through a clothes closet?). What I hope to see some day is an operating system which is smart about speeding up hard drive operations. Solid state drives are going to help but we need something like the DAO “trick” to stay at warp speed (RAM) as much as possible instead of dropping to impulse (slow hard drive operations). Comments welcome.

# 12 brian t said, on January 9th, 2009 at 12:31 am

Since it’s going to be RAID-0, can we assume that this array will not be holding the primary copy of the database, or that there will be regular backups? Striping with no redundancy = one drive drops out, your database is gone, and even if you get the drive back, the database could be corrupted. RAID-0 is no place for critical data!

# 13 OmegaMan said, on January 9th, 2009 at 2:38 am

HD Tune Marc…give us some bench marks. Have Dvorak pony up the money for the full read/write test software.

Try one solo first, then raid them, then test.

# 14 KD Martin said, on January 9th, 2009 at 5:19 am

#12, “RAID 0 is no place for critical data.” Especially when it’s used with that buggy MySQL.

I still favor hot-swap mirror backup, and without multi-raid implementation (ie., RAID 1+0) that’s not going to happen.

# 15 sam said, on January 9th, 2009 at 5:39 am

Google “vista thrashes hard drive”. I bought a desktop with vista basic it thrashed the hard drive constantly. I installed Debian no more problem. I have a labtop with vista premium and I do not notice the thrashing. I prefer a 1tb hard drive for $199 than a 30gb ssd for $75.
SSD could be very bad for the future development of standard Hard drives. If you run amd64 debian with 4gb or more of ram no need for extra speed of ssd over greater storage space of standard 1tb sata.

# 16 sam said, on January 9th, 2009 at 5:45 am

Scratch that I just checked computer deli price list.

5973/I SATA 0250G 7200 WD KS $70.00
6231/I SATA 0320G 7200 WD KS $77.00
7231/I SATA 0500G 7200 Sea $79.00
6308/I SATA 0500G 7200 WD KS $87.00
7767/I SATA 0640G 7200 WD KS $93.00
7366/I SATA 0750G 7200 32 Se $115.00
7436/I SATA 1000G 7200 32 Se $134.00
7765/I SATA 1500 7200 Sea $174.00

# 17 Somebody_Else said, on January 9th, 2009 at 5:55 am

#8
SSDs are flash disks and use the exact same type of memory, they just have a much higher capacity. Thus they have more space to spread writes out to for better wear leveling.

Are they more reliable than regular drives? Nobody knows yet, they haven’t been in the wild long enough. They are certainly much more resistant to shock/heat/etc since they have no moving parts.

#14
Vista spends a lot of time reading/writing for awhile after you initially install. It does a lot of indexing and such.

Turning off/moving the pagefile and turning off drive indexing can minimize unnecessary writes to the SSD.

# 18 QB said, on January 9th, 2009 at 5:59 am

I’m not sure you’d want to put a whole db onto it but certain parts like tempdb make sense. Some db vendors are also experimenting with their databases so commonly read tables/data are mirrored between SSD and hard media.

#10 You’re right. Some of these comments are good, others are straight copy and pastes from Wikipedia. Occasionally it’s fun to Google these comments to see who’s cheating. Right Rob?

# 19 Alex Wollangk said, on January 9th, 2009 at 5:59 am

#14: $199 for a 1TB? Newegg has the new Seagate Barracuda 1.5TB for $129.99 with free shipping…

What I’d do, personally, is set up a raid 0 boot drive with the OS and applications and then put data and swap files on a standard spinning drive with a nice big cache. You get the best of both worlds that way. Add a nice fast USB stick for ready boost in Vista and you have a pretty speedy system.

One thing that might be interesting is a really short adapter that goes from one of the standard USB header connectors to a female USB. Then an enthusiast can plug a USB thumb drive in internally to use for ready boost… I never did like having things hanging on the outside of my computer if I don’t have to…

# 20 Rob Alter said, on January 9th, 2009 at 6:02 am

SSD’s can be Flash-based or DRAM based.

DRAM based have very fast write times. Flash based are slow to write.

I believe someone mentioned these drives slow down on writes so they are probably Flash-based drives.

Flash based drives have limited write cycles. DRAM based do not have this problem.

Flash based drives as a RAID in a server situation should be interesting…

# 21 Somebody_Else said, on January 9th, 2009 at 6:25 am

#19
Yes, the SSD’s we’re talking about here are flash based.

DRAM is very fast and has unlimited writes, but its volatile. You have to have constant power and the cost of purchasing RAMdrives+sticks of RAM to populate them prohibitive.

# 22 Mr Diesel said, on January 9th, 2009 at 9:52 am

Holy cow!! A real technology post on this blog.

Good info for a change…….

# 23 Dan Bran said, on January 9th, 2009 at 10:16 am

I don’t understand the reason for raid 0 here. Raid 0 stripes multiple drives to reduce seek (head movement) times. Since there are no physical moving heads on an SSD it should make no difference. Unless all you are after is a single 120GB solid state drive that is 4 times more likely to fail than any of the individual 30GB drives

# 24 Rich said, on January 9th, 2009 at 10:26 am

“I don’t understand the reason for raid 0 here. Raid 0 stripes multiple drives to reduce seek (head movement) times. Since there are no physical moving heads on an SSD it should make no difference.”

I believe distributing data to more than one drive at a time would improve speed regardless of the technology involved.

# 25 The0ne said, on January 9th, 2009 at 11:06 am

I work with MTBF numbers every month. They should be taken lightly as they only represent the population and can be highly skewed one way or the other. MTBF is not a very accurate but it gives you an overall “feeling” of how your product or part will be doing.

# 26 Kevinb said, on January 9th, 2009 at 12:04 pm

@Alex Wollangk

For a laptop you can simply get a PCMCIA or express card CF slot for a few dollars and keep a flash card internally(and you can find fast CF cards cheap), and for a desktop search Newegg for a “5 port USB PCI Card”. These cards have 4 external and 1 internal USB 2.0 slot, which you can then use to attach your flash stick. I picked one up at Newegg last week for $10.50 including shipping. So while I don’t use Vista the internal boost you desire is quite easy to accomplish. Enjoy!

# 27 ECA said, on January 9th, 2009 at 1:04 pm

sETTING A RAID 1-2 would be nice, and would spread out the MTBF..
grouping the bandpass should accelerate the transfer also..
Im wondering how the mobo is dealing with it..Is it maxing out?
By what I have seen, the amount of power used in this config, is about the same as a Hard drive of the same size. so not allot of real savings.

# 28 bobh said, on January 11th, 2009 at 6:46 am

I put an SSD in a previous generation MacBook Pro. The performance increase is amazing, especially compared to the stock 5400 rpm drive.

Details at:

http://osx411.com/installing-a-128gb-ssd-into-an-old-macbook-pro/

Bob

# 29 Jacques Auef said, on January 11th, 2009 at 10:15 pm

Uh John, you should have called Leo on his radio program and asked him for help!

# 30 Somebody_Else said, on January 11th, 2009 at 10:15 pm

Hmmm,what RAID controller are you using? What speeds are you actually getting?

I can imagine that writes would be slow but you should be getting great reads, although I haven’t used OCZ’s drives myself.

I’ll take the drives if you don’t want them…

# 31 Jim said, on January 11th, 2009 at 10:57 pm

Well from what I’ve been reading the buffered write rates are in the 150MB/s range, so 200 seconds for one drive on average – I presume you’ve already raid-ed them in which case you may be getting better than that actually. I would have to say formatting isn’t a good identifier for benchmarking though, it’s a constant write stream with very little reading.

Make sure that you are NOT putting any paging or heavy writing filesystems on it though. The random write/reads will ruin whatever possible performance gains you’ll get. I’d use it for OS system files and distributions and databases where the largest part of the system is read-only.

Now it would be interesting to have a secondary SATA hard drive holding your high TPS database files, with the ssds holding all of your lookups and low-write tables. I suspect it would be on par with caching it all in memory, though you still have the overhead of the SATA controller access.

Did you have some particular reason to strip it instead of going for a raid-5 config? With new drive types it’s far more realistic (and comfortable) to set them up as raid 5 so you can potentially hot-swap out a bad drive and not lose the whole array.

You actually should be able to use that raid box that John was pushing a few months back, I can’t remember but I thought it could handle raid-5.

# 32 Marc Perkel said, on January 11th, 2009 at 11:16 pm

I’m using Linux software raid and the motherboard has 4 SATA II ports. the drive is rated at 150/90 r/w speed which is reasonable fast. In Raid it should be about 4X that speed. What using real hard drives this works. So what I suspect is that those numbers are cooked and that in real life you don’t get that speed under load with random reads and writes.

# 33 cgp said, on January 12th, 2009 at 12:05 am

Check out that macbook 128gb ssd upgrade.

After a while the mac crawled to a halt … defective blocks,
ssd returned on warranty.

# 34 Somebody_Else said, on January 12th, 2009 at 12:19 am

HD-Tach shows my 64 GB G.Skill drive is reading just below 150 MB/s with an access time of .2ms. Linear writes like copying a file to the drive gives me ~30-50 MB/s. The documentation says that I should be getting 155 MB/s max on reads and 90 MB/s max on writes. Doing a NTFS format (not a quick format) of the drive in Windows setup was near instantaneous.

What read/write performance you’re actually getting? Have you tried testing the drives individually to make sure you don’t have a bad one? That’s all I can think of.

I wouldn’t expect 4x performance since the seek times are already so low on SSD’s, but it should help with write performance.

# 35 KD Martin said, on January 12th, 2009 at 12:31 am

Marc, what speed testing diagnostics have you tried? SI-Soft makes a good one, and there are several others. I’m not certain how they’ll work on SSDs, but it might be worth a try.

# 36 Chris Mac said, on January 12th, 2009 at 1:23 am

Are the raid drivers spec’d to perform this task?

# 37 amodedoma said, on January 12th, 2009 at 2:50 am

the bios of the raid/sata controller should be accessible, did the drives auto detect ok?

# 38 QB said, on January 12th, 2009 at 5:57 am

SSD has a completely different “behavioral topography” from hard media. If the OS and drivers aren’t designed for that then it’s inefficient.

A combo of hard and SSD can work well, and will increasingly get better. Unfortunately, the Linux folks are well behind on this one and need to catch up.

# 39 Skaiye said, on January 12th, 2009 at 6:33 am

Raid controllers use memory buffers to speed up the transfer rates on regular hard disks and Raid controllers can actually hinder SSD performace. Unless you get a specific type of Raid Controller for SSD performance then you wouldn’t see a performance increase you will actually see a decrease.

# 40 Marc Perkel said, on January 12th, 2009 at 7:17 am

Why should it take a specific driver for SSD? It’s a SATA II drive. It should act like any other drive.

As to what speed test am I using – none. It’s just that a format process that should take 10 seconds is taking 5 minutes.

# 41 Mark Derail said, on January 12th, 2009 at 7:47 am

Raid performance off a motherboard, your speed will be slow.

Worth spending the extra $ for a dedicated PCI-Sata-Raid card that supports four HD’s.

Then again, nothing beats older SCSI hardware. Sub 36G’s SCSI-2 are usually below 20$ (used) and servers with a six-drive SCSI backplane, easily found for sub 400$.

Database performance on such machines is high, however, also a high HD replacement rate.

Depends if you “see” your machine on a daily basis or not.

# 42 Marc Perkel said, on January 12th, 2009 at 8:11 am

I’m running several other raid arrays, Raid 0, Raid 1, and Raid 10, all off the motherboard using software raid and they are all fine and fast. The fancy raid controllers only help on raid 5 and 6.

# 43 BigBoyBC said, on January 12th, 2009 at 8:23 am

Sorry Marc, I was hoping this would have worked, I guess that’s why the drives are so cheap.

Keep trying…

# 44 FRAGaLOT said, on January 12th, 2009 at 9:14 am

#11

Very intresting, but back in the old DOS/486 days we weren’t using an OS that used a “virtual memory” file that would cause heavy drive access. Ever since windows 3.1 came around there was this use of virtual ram since for some reason we needed more RAM than we actually had.

20 years later we are using systems that have at least 2-4 GIGABYTES of RAM, which larger than the first set of hard drives I had in 10 years ago.

Yet why are we still using virtual ram? Do we really still need it now? Why would simple apps like web browsers, Email and IM software need a giant swap file of 8 gigs?

# 45 Somebody_Else said, on January 12th, 2009 at 12:26 pm

“# 42 Marc Perkel said, on January 12th, 2009 at 8:11 am

I’m running several other raid arrays, Raid 0, Raid 1, and Raid 10, all off the motherboard using software raid and they are all fine and fast. The fancy raid controllers only help on raid 5 and 6.”

Now I’m confused. Are you using Linux “softRAID” or your motherboard’s “fake RAID.” Both are software RAID and you may see bad SSD performance with some of the older motherboard RAID controllers.

I would try using the motherboard RAID and install Windows or some other OS to see if its just a Linux problem. Hook the drives up individually to another machine, format them, and run a speed test to make sure you don’t have a bad drive in the mix.

# 46 Marc Perkel said, on January 12th, 2009 at 12:28 pm

If I had a bad drive it wouldn’t work at all.

# 47 Somebody_Else said, on January 12th, 2009 at 1:49 pm

Not necessarily, you might have a bad drive that’s getting crappy read/write performance and slowing the whole array down.

It’s not like its that hard to check. Just hook them up to a Windows machine and run a free benchmark like HD-Tach on each drive and see what it says.

# 48 Phace said, on January 12th, 2009 at 5:33 pm

OCZ’s Core drives suck. I could have told you that. I have a 120gb model. They freeze for a good 10 seconds sometimes, and are slow for anything but a pure read.

Now… I did get it to work quickly- Windows Disk Protection- It caches writes to a log file and commits them at start-up. Makes the drive fly. It’s the random writes that kill these SSD’s.

# 49 deowll said, on January 12th, 2009 at 6:42 pm

I’m don’t recall why, I did read it, but my flash drives are much slower reading than writing. They are slower than the hard drive. I have noted that if I stick a lot of pictures in a file then copy the file for some reason that is much faster than just copying the individual pictures.

I have a chip reader and bought a large size chip (4 gig but the compactFlash I/II/MD itself is what is over sized) to speed up boot time on my Vista machine. It works on my HP.

At a guess my machine mainly boots off the chip and has a few read files there.

Considering that I like to save my working files often in case of failure and office does temp files and flash memory is slow to write and is reported to be failure prone due to being written too to oftern I don’t want my data storage file to be flash. Hard drive space is cheaper and faster. I have 28 gig of pictures and am starting to do video.

My take on built in fash drives is stick your OS and programs on the flash drive and your data on a hard drive/drives.

Leave Your Reply Below...

Comment Moderation is Active, so if your comment doesn't
show up immediately, please don't submit it again.

Special Deals 4U

25% off eHarmony subscription with code EHTECH
12-percent discount forever with code TECH
10-percent discount and free month using TECH10
10-percent or 30 dollars off weekly car rental!
Get 10-percent discount on car rental!
Get 10% off rental cars!
Get 10% off Dollar rentals!
Discounts on domain names!