Poor LSI SAS1068E Write Performance with Linux
While doing research into poor write performance with Oracle I discovered that the server was using the LSI SAS1068E. We had a RAID1 setup with 300GB 10K RPM SAS drives. Google provided some possible insight into why we the write performance was so bad(1 2). The main problem with this card is that there is no battery backed write cache. This means that the write-cache is disabled by default. I was able to turn on the write cache using the LSI utility.
This change however did not seem to any difference on performance. At this point I came to the conclusion that the card itself is the blame. I believe that this is an inexpensive RAID card that is good for general use of RAID0 and Raid1, however for anything were write throughput is important, it might be better the spring for a something a little bit more expensive.
When it was all said and done we ended up replacing all the these LSI cards with Dell Perc 6i cards. These cards did come battery backed…which allowed us to then enable the write cache, needless to say the performance improved significantly.
I am currently working as a Sr. Linux System Administrator for National Public Radio in Washington D.C.
ur | February 4th, 2010 at 10:31 am #
Hey, I’ve got the same LSI SAS controller and i’m thinking about replacing it with Dell Perc. Can you tell me, how big is the performance boost? Have you got any benchmarks?
shainmiley | February 4th, 2010 at 11:16 am #
Well I don’t have any hard benchmarks to show you, however since we were running Promox on these servers we were using a perl script that comes with that distro as our benchmarking tool.
The script is called ‘pveperf’. Here is the output from the command:
Original LSI SAS1068E output:
BUFFERED READS: 115.99 MB/sec
AVERAGE SEEK TIME: 6.58 ms
FSYNCS/SECOND: 272.47
New Dell Perc output:
BUFFERED READS: 101.11 MB/sec
AVERAGE SEEK TIME: 6.33 ms
FSYNCS/SECOND: 2509.68
Clearly the seek time should still be about the same…as should the buffered reads. The major performance gains are seen in terms of the write speed (FSYNCS/SECOND) as there is almost a 10X increasing using this metric.
I can tell you for a fact that if you have an I/O intensive application (Mysql, Oracle, Samba/NFS server with heavy writes) for example…you will see a huge increase in performance….and it is all due to the ability of the controller to cache those writes.
We ended up replacing these LSI cards with Perc ones in 10 to 15 of our Dell PowerEdge servers after this discovery.