Unsolved
This post is more than 5 years old
306 Posts
0
847
How to interpret what I am seeing in Performance Detail.
We have an application that has a MetaLUN.
This MetaLUN is spread across 10 RAID 5 groups (4+1). The DB guys ask if we are seeing any performance issues with their LUN.
I open Performance view in Analyzer and select the LUN and the disks. When I look at response time I have a spike that reached 16million.. yes MILLION milliseconds for the LUN. All the disks (All 50 of them) are less than 40ms response time.
So my question is... Why would I see such a large response time viewing the MetaLUN and see good response times from the disks which make up that MetaLUN?
Thanks!
This MetaLUN is spread across 10 RAID 5 groups (4+1). The DB guys ask if we are seeing any performance issues with their LUN.
I open Performance view in Analyzer and select the LUN and the disks. When I look at response time I have a spike that reached 16million.. yes MILLION milliseconds for the LUN. All the disks (All 50 of them) are less than 40ms response time.
So my question is... Why would I see such a large response time viewing the MetaLUN and see good response times from the disks which make up that MetaLUN?
Thanks!
DGM3
238 Posts
0
February 18th, 2008 10:00
A prior thread that briefly discussed the same, is here;
http://forums.emc.com/forums/thread.jspa?threadID=58987
Regards,
DGM
DHoffman2
306 Posts
0
February 18th, 2008 11:00
CX3-80 FLARE Code 24 Rev. 11
The SP does indicate a ver low response time, except for some peaks that go as high as 224ms. Good follow up question.
I'll review the link you sent DGM and see if that shed any additional light.
DGM3
238 Posts
1
February 18th, 2008 11:00
If you may want to open a support case (if logs showing the event in question can be available for review) the support team could likely offer further comment. If you do open a case, please advise back here what the SR number assigned is.
Regards,
DGM
DHoffman2
306 Posts
0
February 18th, 2008 11:00
This is what I am seeing, however do you know if EMC plans on resolving this "false positive"?