zhihui,
Filesize is only used to determine how large to make the file
initially. If you do sequential writes to it, as those two workloads do,
it will continue to grow.
zhihui Chen wrote:
I believe we have an RFC to add an attribute such as "maxfilesize" to
filesets, which will prevent them from increasing beyond that size. You
could set filesize to one size, then maxfilesize to another to set the
maximum size your file could grow to due to appends and sequential
writes beyond the current end of file.
We are not sure what should happen when they hit that limit though.
Two possibilities are: 1) start over from the beginning of the file, 2)
just quit working.
There is also the question of how to set this limit. With a single
file you can just set a hard limit, but with a fileset you would want a
different limit for each file, related to the files initial size. One
possibility would be to allow you to set a relative limit as a parameter
to the fileset, such as maxfilesize=120%.
Do you have any preferences regarding the two issues I raised above?
Drew
I set filesize to 2048m for singstreamwrite and multistreamwrite,
which I think that filebench will generat 2048m-size file. But
during filebench running, I find size of target file is increaing
consistently until ending of filebench, which leads to 30GB+ target
file. Running time for these two workloads are set to 60 seconds
and the tested storage can generate throughput 480+MB/s. Does this
mean that filesize parameter for these two workloads is meaningless?
Thanks
Sam
--
------------------------------------------------------------------------
_______________________________________________
perf-discuss mailing list
perf-discuss@opensolaris.org
_______________________________________________
perf-discuss mailing list
perf-discuss@opensolaris.org
