disk performance

disk performance

Post by Ume dandike » Sat, 17 Oct 1998 04:00:00



Is their any utility/command I can use
to determine if there is a disk performance
problem.   I did notice a high
disk service time in the iostat report?

No error messages reported by root and
I have now turned on all debugging logs.

Any suggestions to rule out disk error/
performance???

Thnx in advance

Ume

 
 
 

disk performance

Post by sami » Sat, 17 Oct 1998 04:00:00


Ume,
Performance depends on the Disk configuration(RAID ,SSA etc) and the
SCSI interfaces and the Model of the machine(Iam sure you know that).
Just iostat will not be enough to trace the disk performance issue .If
you see too much pagein page out in vmstat then you might have to
distribute your Paging space acroos VG or PV's.USe vmstat,iostat and SAR
commands in sync and try to get a feel of which part of Disk is causing
the performance bottleneck.
It is advisable to keep your jfs logs(which is accessed the most ) in a
different disk then some of your application(maybe say Oracle ).So
identify the bottleneck and distribute your applications.
-samit

> Is their any utility/command I can use
> to determine if there is a disk performance
> problem.   I did notice a high
> disk service time in the iostat report?

> No error messages reported by root and
> I have now turned on all debugging logs.

> Any suggestions to rule out disk error/
> performance???

> Thnx in advance

> Ume


 
 
 

disk performance

Post by AIXMight » Thu, 22 Oct 1998 04:00:00


Ume,



>Is their any utility/command I can use
>to determine if there is a disk performance
>problem.   I did notice a high
>disk service time in the iostat report?

depending on the oslevel you have, you can check /usr/sbin/perf/,
/usr/lpp/*perf*, filemon, fileplace and so on to help you gathering some
stats...

Quote:>Any suggestions to rule out disk error/
>performance???

yea but i'd better refer you to Frank Waters' Performance Tuning Guide.
or you can try the online one at
http://www.rs6000.ibm.com/doc_link/en_US/a_doc_lib/aixbman/prftungd/t...

Orphy! :-)

 
 
 

1. Multi disk performance (8 disks), limit 230 MB/s

Hi all,

I have been doing some benchmarking experiments on kernel 2.4.19 with 8
IDE disks. Due to poor performance of 2 disks at 1 IDE channels we have
bought 4 Promise Ultra 100 TX2 (32-bit 66 Mhz) controllers and to avoid
bus saturation Supermicro P4PDE motherboard with multiple PCI buses
(64-bit 66 Mhz) and 2-Xeons. I submitted already PCI slot placing
problems to the mailing list. But theoretically I can live with the
current IDE condrollers->PCI slots assighnment.

The assignment is the following: 3 IDE controllers are connected to the
one PCI 64-bit Hub with bandwidth 1 GByte/s and  4th controller is on
another hub with the same characteristics.

Theoretically with 6 IBM disks (47 MB/s from the first cylinders) I
should achieve a number about  266 MB/s (32 bit X 66 Mhz) < 6*47. AND
2*47 = 94 MB/s < 266 MB/s from the last two disks. Thus the rate should
be 94 + 266 = 360 MB/s.

BUT no matter from which set of the disks I read or write I have got the

following parallel read/write rates (raw access):

         write (MB/s) read (MB/s) systime (top)  real/user/sys(time) (s)

1 disk :        48    45          3  %           3.0 / 0.1 / 0.4
2 disks:        83    94          10 %           3.5 / 0.1 / 0.6
4 disks:        131   189         21 %           4.3 / 0.4 / 2.8
5 disks:        172   233                        4.5 / 0.5 / 4.5
6 disks:        197   234 ?       30 %           5.2 / 0.6 / 6.6
7 disks:        209 ? 230 ?                      5.9 / 0.6 / 8.8
8 disks:        214 ? 229 ?       40 %           6.7 / 0.8 /10.8

The method of the measurement:
1. Prespawn thread for every disk
2. signal threads to start write of 64 Mb of data
3. wait until all threads finished, measure time
4. signal threads to start read of 64 Mb of data
5. wait until all threads finished, measure time

All disks are in UDMA 100 mode

Has anyone else seen similar problems?
What limits the performance: IDE disk driver, bottleneck in the kernel
I/O subsystem?
How to improve?

Roman

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in

More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

2. FCA-PCI question

3. Poor disk performance sparc20 2.4

4. X Client Command - Help

5. Disk Performance Problem

6. how to pass values from lilo's menu to /etc/rc.d/rc.local

7. IDE disk performance

8. @home,rh6.0,3com el3c589 woes

9. disk performance on ASUS P54NP4D motherboard

10. Tuning disk performance in FreeBSD?

11. Disk performance on SS20 w/320MB running Solaris 2.3 ...

12. Solaris 8 slow disk performance with journaling enabled

13. very slow disk performance..- 3.3-RELEASE