Re: Testing Sandforce SSD - Mailing list pgsql-performance

From Yeb Havinga
Subject Re: Testing Sandforce SSD
Date
Msg-id 4C4B5984.4010705@gmail.com
Whole thread Raw
In response to Re: Testing Sandforce SSD  (Yeb Havinga <yebhavinga@gmail.com>)
Responses Re: Testing Sandforce SSD  (Greg Smith <greg@2ndquadrant.com>)
List pgsql-performance
Yeb Havinga wrote:
> Yeb Havinga wrote:
>> diskchecker: running 37 sec, 4.47% coverage of 500 MB (1468 writes;
>> 39/s)
>> Total errors: 0
>>
>> :-)
> OTOH, I now notice the 39 write /s .. If that means ~ 39 tps... bummer.
When playing with it a bit more, I couldn't get the test_file to be
created in the right place on the test system. It turns out I had the
diskchecker config switched and 39 write/s was the speed of the
not-rebooted system, sorry.

I did several diskchecker.pl tests this time with the testfile on the
SSD, none of the tests have returned an error :-)

Writes/s start low but quickly converge to a number in the range of 1200
to 1800. The writes diskchecker does are 16kB writes. Making this 4kB
writes does not increase writes/s. 32kB seems a little less, 64kB is
about two third of initial writes/s and 128kB is half.

So no BBU speeds here for writes, but still ~ factor 10 improvement of
iops for a rotating SATA disk.

regards,
Yeb Havinga

PS: hdparm showed write cache was on. I did tests with both ext2 and
xfs, where xfs tests I did with both barrier and nobarrier.


pgsql-performance by date:

Previous
From: Greg Smith
Date:
Subject: Re: Testing Sandforce SSD
Next
From: Greg Smith
Date:
Subject: Re: Testing Sandforce SSD