Re: JetStress Available Mbytes

The three numbers are average, minimum, and maximum, so I wouldn't call it a
leak. The average and maximum are higher that the 2.000 minimum, inferring
that memory was allocated and then released. Did you use the max cache
tuning parameter, or were parameters set using autotuning? What paramaters
were used?




"DLane" <DLane@xxxxxxxxxxxxxxxxxxxxxxxxx> wrote in message
news:A88573CA-5B61-4CC4-BFCA-325762FC90A1@xxxxxxxxxxxxxxxx
John,

Thanks for the reply - that doesn't tell me why JetStress causes memory to
go below 2MB - here is the overview of the results - The achieved I/O was
283:

Total test database size Production data size Total number of
databases Expected I/O Achieved I/O
19.49 GB (based on the attached database) (n/a) 6 (1 storage(s) * 6
database(s)) 100.00 (200 mailboxes of 0.50 IOPS) 283.00

JetStress test parameter summary:
Instance threadCount logBufferSize opInsert opReplace opDelete lazyCommit
1 1 9000 26 63 2 89

Disk subsystem performance summary:
Volume Avg. Disk sec/Read Avg. Disk sec/Write Disk Reads/sec Disk
Writes/sec Avg. Disk Bytes/Write
Data (F:) 0.005 0.004 162.19 120.82 (n/a)
Log (G:) 0.000 0.002 0.011 32.39 6693.72

Processor/memory performance summary:
Counter Average Minimum Maximum
% Processor Time 0.912 0.156 2.077
Available MBytes 118.17 2.000 749.00
Free System Page Table Entries 252253.96 252235.00 252254.00
Pages/Sec 60.89 0.000 1836.48
Pool Nonpaged Bytes 27632843.03 27619328.00 27697152.00
Pool Paged Bytes 22262969.38 21921792.00 22315008.00
Database Page Fault Stalls/sec 0.000 0.000 0.000

The disk result seems accurate - I have a 15 spindle, 10k RPM Fibre SAN
using virtualization to use all 15 spindles for each volume - RAID 10 and
RAID 5-5 combined to provide top performance. My main question is the
memory
result. That is the only failure I am encountering and cannot find
anything
in the perfmon results pointing to anything other than a memory leak in
Jetstress.

I may just call Microsoft, but was hoping others had seen a problem like
this.

"John Fullbright [MVP]" wrote:

Jetstress isn't all that granular, so you should be a bit careful when
interpreting the results. Compare the expected IO (derived form the
iops/mailbox inputs and number of mailboxes) against the achieved IO
(based
on the tunings and thread count the IO level that was actually achieved,
and
presumably the basis for the rest of the measurements). It looks like
your
expected IO would be in the 100 IOPS range while your achieved would -
with
one thread - be in the 600 IOPS range; not eactly appropriate relative to
the specific disk subsystem. See jetstress.doc page 33.


"DLane" <DLane@xxxxxxxxxxxxxxxxxxxxxxxxx> wrote in message
news:517AE5F7-FE2A-41B4-9186-0FA0A66EE734@xxxxxxxxxxxxxxxx
Hello,

I have the latest version of JetStress - 6/7/2006 - and am using
Exchange
2003 SP2's ese.dll and other ese files to perform testing.

The server is a Dell 1425 with 1GB of RAM and local drive with the page
file
configured. I am connecting to a fibre SAN and have set the /3GB swith
along
with the /USERVA=3030 in addition to the registry entries required
according
the optimizing memory article for Exchange 2003.

I am using the default settings for JetStress when running the 2 hour
performance test. I have setup the system with 2 seperate volumes -
one
for
the databases and one for the logs - I am using 1 storage group and
configured the test for 100MB mailbox limit, 200 users and .5IOPS per
user.

The storage solution is return disk sec/read and writes in the area of
.005
with a high of .008 - no disk queueing, etc.

The problem is the Available Mbytes ALWAYS drops below 2. I have tried
two
different servers and the only problem I can see in the Jetstress runs
is
that Available Mbytes drops below the acceptable level - the behavior
seems
to be the same regardless of the /3GB setting - it looks like a memory
leak.
Anyone else seeing this?





.

Relevant Pages

  • Proposed Assembler Commands
    ... ACM Automatically Clear Memory ... BKCRDR Backspace Card Reader ... BKSPD Backspace Disk ... EIAO Execute In Any Order ...
    (sci.electronics.design)
  • Re: Att. Alex Nichol -VM cont.
    ... I find documentation that says a 4kb page in memory is written to the hard ... manner that a 4kb memory page is equal to a 4kb cluster written on the disk". ... So where is the basis for saying since 4kb paging in memory that 4kb clusters ...
    (microsoft.public.windowsxp.general)
  • Re: Slow performance
    ... Leaving your computer on 24/7 means that if any programme has a memory ... The Disk Defragmenter report shows you have minimal free disk space. ... select Properties, General, Advanced and check the box before Compress ... Volume fragmentation ...
    (microsoft.public.windowsxp.perform_maintain)
  • Re: Slow performance
    ... I realize that the pagefile usage is excessive and I need to reduce memory ... I've never messed with those settings before. ... Peak are greatly in excess of the installed RAM. ... The Disk Defragmenter report shows you have minimal free disk space. ...
    (microsoft.public.windowsxp.perform_maintain)
  • Re: teaching a child - console or GUI
    ... I use a RDBMS to reuse existing code and optimizations. ... >>And loading them in a system under load (with constant disk io) is worse ... quality stuff), and quality power supply, quality memory etc. ... > lists of pointers on disk, extract files of frequent search fields in ...
    (comp.lang.pascal.delphi.misc)