This post is more than 5 years old
2 Intern
•
360 Posts
0
4547
March 29th, 2008 11:00
About "Average Busy Queue Length"
Dear All,
I would like to check the performance of specific RAID Group is good or not.
Can I check the value of "Average Busy Queue Length" to verify?
what is the good value if I can use it to check?
Thanks for any reply.
Dennis
I would like to check the performance of specific RAID Group is good or not.
Can I check the value of "Average Busy Queue Length" to verify?
what is the good value if I can use it to check?
Thanks for any reply.
Dennis
No Events found!
kelleg
6 Operator
•
4.5K Posts
0
April 4th, 2008 09:00
regards,
glen
RRR
6 Operator
•
5.7K Posts
0
April 7th, 2008 00:00
sarpydog
2 Intern
•
360 Posts
0
April 12th, 2008 02:00
Is it possible that value are not accurate?
Dirty Cache % are always under 80 and no I/O size is greater than 1023K.
So furthur I/O will fit into cache and then state to physical disk.
is above concept right?
RRR
6 Operator
•
5.7K Posts
0
April 14th, 2008 02:00
kelleg
6 Operator
•
4.5K Posts
1
April 16th, 2008 14:00
If using NAR files to track down your problem, use 60 seconds to get greater detail.
Look at queue length (as measured value) and Total Throughput (another measured value). Look for LUNs with high queue length (over 10), Utilization over 70% and response time over 20ms at the same time point. You need all three of these values to exceed the Best Practices guidelines.
Dirty Pages going to 80% and then dropping back to 60% indicates that the watermarks are operating correctly - keeping your write cache efficient. Over 80% and you start to force flush. If you hit 99%, the write cache shut down until the writes are de-staged to disk. If the disks can't keep up with the writes, the queue increases and write cache dirty pages can stay at 99%. Too many writes or IO's to the LUN can bog down the disks.
glen
sarpydog
2 Intern
•
360 Posts
0
April 17th, 2008 05:00
Thanks for your response.
I will check the archive interval set, I used the default value about it.
The Cache usage are good for normal operating, but I may get higer respone time.
The root cause may be followings:
1. CX3 mirror to CX700 and CX700 is the bottleneck.
2. I can see any details of mirrored LUN via Navisphere Analyzer. Is it normal?
3. CX700 used the temp anlyzer license for troubleshooting.
Best Regards,
Dennis
sarpydog
2 Intern
•
360 Posts
0
April 17th, 2008 05:00
I forgot a word.
I can not see any details of mirrored LUN via Navisphere Analyzer. Is it normal?
sarpydog
2 Intern
•
360 Posts
0
April 17th, 2008 07:00
Thanks for your quickly response.
I can not see any performance data of mirrored LUN on remote storage (CX700).
Is it possible cause that we use Rebootless Analyzer enabler for R14, R16, R17, R19 and R22 on remote storage?
Best Regards,
Dennsi Dai
odurasler
1 Rookie
•
103 Posts
0
February 16th, 2010 09:00
I know this is an old thread, but I wanted to respond to your question just in case you didn't get the answer. I'm sure you already know though.
My understanding was that because the older Clariion was running on a Windows 32bit OS that it cannot have more than 4GB. And the new Clariions are supposed to be running on 64bit OS now.
I could be wrong though.