Le 3 août 09 à 12:06, Scott N. a écrit :
Pretty standard desktop type machine but I threw in an old
3Com-905TX NIC into because the onboard nvidia nic didn't work with
Opensolaris 2009.06. My other machine is running Vista SP1 x32 and
using onboard broadcom nic. Both PC's are on same netwo
The 8K record is what optimises the transactional performance at the
expense of file scan.
Now you could revert to a 32K recordsize if you want a different
tradeoff but also did you consider
using snapshots and zfs send | zfs recv ?
That will allow you to replicate the DB incrementally only c
Moreover, even if you'd find a way, wouldn't that save something < 1
usec of the 1 millisec pagefault ?
A big contortion to get 0.1% improvement. I'm sure engineering time
can be better spent.
Le 24 sept. 07 à 20:55, [EMAIL PROTECTED] a écrit :
>>My means is not prefetch the page but u
I would like to caution people about this conventional
wisdom. For UFS the I/O size tunable like maxphys ended up
as an important parameter to the write throttle algorithm.
The bigger it was the less coupled the application would be
to the storage and some important gains could come from
that.
Sa
ified a problem with the CAS implementation
used in Solaris ?
-r
________
Roch BourbonnaisSun Microsystems, Icnc-Grenoble
Senior Performance Analyst 180, Avenu
Eric Lowe writes:
> On Wed, Jan 25, 2006 at 03:04:55AM +1100, Brendan Gregg wrote:
> |
> | The way a freshly written file (as opposed to a freshly mounted file)
> | changed the number of faults really threw me. Still haven't put my finger
> | on why, but I suspect it's a cache placement poli
In case I wasn't clear, given the current situation we definitively need this
fix-in as soon as possible.
+1
-r
Roch BourbonnaisSun Microsystems, Icnc-Grenoble
Senior Perfor
Codereview comment for
P3 kernel/other cpu_X_system kstats should include interrupt time
NITS: In machdep.c* we could make intracct[] as hrtime_t ?
uint64_t intracct[NCMSTATES];
Otherwise all's well.
-r
___
perf-discuss mailing l
