On May 14, 2008, at 3:24 PM, Robert Milkowski wrote:

> Hello zfs-code,

hiya

>
>
>
>
> http://cvs.opensolaris.org/source/xref/onnv/onnv-gate/usr/src/uts/common/fs/zfs/vdev_cache.c
>     72  * All i/os smaller than zfs_vdev_cache_max will be turned into
>     73  * 1<<zfs_vdev_cache_bshift byte reads by the vdev_cache (aka  
> software
>     74  * track buffer).  At most zfs_vdev_cache_size bytes will be  
> kept in each
>     75  * vdev's vdev_cache.
>
>
> While it is mentioned in lines 46, 47 that it does affect only  
> metadata
> I/O, imho for clarity it should also be mentioned in lines 72-75.

Yeah, "All i/os smaller..." seems to contradict the earlier lines.  We  
can fix that.

>
>
>
> I guess 
> http://cvs.opensolaris.org/source/xref/onnv/onnv-gate/usr/src/uts/common/fs/zfs/zio.c#963
> http://cvs.opensolaris.org/source/xref/onnv/onnv-gate/usr/src/uts/common/fs/zfs/vdev_cache.c#272
> are responsible for assuring that vdev_cache is used only for  
> metadata.

Yep.

>
>
>
> btw: above code changes were introduced as part of
> 6437064 (with further changes)- has it been backported to S10 and/or  
> delivered as patch?

6437054 vdev_cache wises up: increase DB performance by 16%
has already been putback into build 1 of s10u6.

eric

>
>
>
>
>
> -- 
> Best regards,
> Robert Milkowski                          mailto:milek at task.gda.pl
>                                     http://milek.blogspot.com
>
> _______________________________________________
> zfs-code mailing list
> zfs-code at opensolaris.org
> http://mail.opensolaris.org/mailman/listinfo/zfs-code


Reply via email to