MySQL with SSD drives
Hey everyone..I want to get some feedback. I was wondering if anyone had experience with SSD (Solid State Disk) drives. They are basically flash memory mounted and designed to be used as a replacement for a typical hard drive. The prices are really falling on SSDs and, despite people who say otherwise, they are now as reliable, or more reliable, than typical hard drives. MTBF was rated at 2,000,000 hours for one drive and they all seem to have good warranties (I saw five years on one drive).
What makes me drool is not the throughput of a typical drive (which is about the same as a high-end drive) but the .1 ms random-access seek time. Yes, that is POINT one!! As in 1/10th of a millisecond!! I saw one that was advertised to have a seek time of .03 - .1 milli-seconds.
They are still very expensive. But, what if I replaced the hard drive holding my data files with one of these drives. Now my seek time for reading tables off disk is essentially zero. I wonder how that would affect performance. Ideally I would like three drives, one for the OS, one for the log files, and one for the data. I bet your I/O bus bandwith/speed would suddenly become your new bottleneck - not your hard drives themselves.
Size of the drive is an issue. Sixty-four GB size drives are now common and 128GB are becoming available. I have several databases that would not fit on the current drives, but still I think I could manage to split things up if needed. Once 128GB becomes the norm even this wouldn’t be a problem.
So, if anyone has had any experience with these I would love to hear from you. Benchmarks would be great!! Now if I could just get a half dozen to throw in my Coraid and test it out..
5 Comments so far
Leave a reply
Just be careful with some of the vendors. They’re lying about their IOPS. They’re claiming higher IOPS than the drive can handle. Some of these disks can only write 1/10th of their read rate.
Kevin
Also be carefull about random WRITE speed. Random read speeds are excelent since it’s basically just memory. But flash has to be block erased, and that means a small random write takes alot of time to perform.
I haven’t checked out the drives recently, but maybe they have overcome this limitation, or just made the block size smaller, and dealed with the issues of doing that also.
As I dig deeper into this I am finding out that the random small file write speed is certainly an issue. There is work to overcome it, and I think over time this will improve to an acceptable level. I would be curious if, in a database environment, the random write issue is not so much of a problem. Most databases that are online are far more reads than writes and the read access time of these “drives” just blows me away.
I have decided that I am going to re-visit this in about six months. If there is promise for better random write speed I can probably get my company to buy me a couple to play with.
I have heard — very unsubstantiated — rumors that some large “2.0″ companies are testing these drives on DB’s or even possibly using them in some fashion in production. I would be fascinated to hear of their experiences if that is true.
Have a look at the Hyperdrive4 from Attorn. It’s RAM based but for $5000 you get something much faster than any hard drive of flash based SSD. It also writes as fast as it reads. Would seem the obvious route for OLTP.
I will take a look at the Hyperdrive4. I have researched the RAM drives a bit but so far they are so ridiculously overpriced it isn’t worth looking at. The idea is certainly attractive.
Thanks for the information!!