Start a Conversation

Unsolved

This post is more than 5 years old

1011

January 29th, 2008 07:00

Can not collect performance data for Connectivity Devices

Hello All,

This is an ongoing issue since one month and until yet EMC support could not find the cause. Opening this question, if anyone of you guys have anythoughts of might be it.

We have a ECC 6.0 and 2 Mcdata ED-10000M switches in our environment. I installed the fiberzonebridge on the Connectrix Manager 9.0 server and was able to discover both the fabrics in the ECC, the performance and WLA are not being collected. we have two FCC agents hosts running one primary for each fabric. After sometroubleshooting, EMC came up with upgrading connectrix manager to 9.5. I did this as well last week and rediscovered the fabric's again on the ECC 6.0. But It's the same issue still. No performance data and WLAs. As far as I think, this is not the issue with switch side i believe. something with the FCC agents or the ECC itself. The connectrix manager has two nic cards one going to both the fabrics and the other one the management network to talk to ECC and our workstations. We can see the stats in the connectrix manager but not being sent to ECC. I have all the snmp setting correctly configured for both CM9.5 & ECC6.0.

Let me know for any questions and any help would be really appreciated as this been a long time, still not able to fix it.

8 Posts

January 29th, 2008 15:00

First the obvious. DCP Fibre Channel Connectivity Agent ->WLA Daily enabled and Apply To tab shows your switch devices?

Second not so obvious. In Performance Manager - Reports -> Job Scheduler. Open up +Analysis-> +Datasets, should see Connectivity Device - All. If not it is possible to restore the original job file under Reports pulldown.

131 Posts

January 30th, 2008 08:00

Yes the fabric's show in the WLA DCP and I did applied to it. As for the performance manager, acutally the stats are not being collected at all. It error out daily iin the ECC saying stats collection failed.

Also I went through some emc primus solutions and found that if there are any unmanaged switches in the connectivity devices the whole fabric rediscovery fails. is this true? in my fabric there are other san switches which are not fully discovered. The connectrix manager sees those switches but did not fully discover as it was populated with ip addresses.

Any idea's???

2 Intern

 • 

5.7K Posts

January 31st, 2008 01:00

Noooooo, just the discovery of that 1 switch fails, but since zoning information is in all switches, in fact you only need to discover 1 switch to get all wwns and information you might need for use in ECC.

But how to solve your problem to get the failed discoery back to working again... beats me. I had the same problem and never got to it... The DCP simply didn't let me re-add the switch for which discovery didn't work anymore. I'm sure this could be a "simple" delete of the fabrics (after disabling the FCC agents) and rediscovery again of everything or a hack in some ini file or so.....

131 Posts

February 1st, 2008 13:00

After some changes and troubleshooting DCP policies for rediscovery is working now. Actually I went and removed the lowest wwn in the fabric on Connectrix Manager which was unmanged and was causing the issue with rediscovery.

But still the WLA and performance stats are failling. Right now I reduced to this one single error in the fcc log files. The error says "ESN API returned 0 handles for ip <192.168.101.40>" and the fails to collect stats.

Any one has any ideas??????

"SDK 15:45:05 C P T 1860 (20.50 MB 64.000000K) ThreadRun => SANITY: Begin internal thread wrapper
SDK 15:45:05 C P T 1860 (20.51 MB 8.000000K) ThreadRun => Start Thread "DataPolicy Thread, Checking
- Performance Statistics", Act: 17, Blkd: 0 (1860)
EGN 15:45:06 C P T 1860 (20.51 MB) [Process DP] => Calling agent to process data policy Performance
- Statistics
EGN 15:45:06 C P F 1860 (20.51 MB 4.000000K) [Process] => SwitchRTPerformancePolicy.cxx@152 : L3:
- Switch WWN list has 2 elements.
EGN 15:45:06 C P T 1860 (20.51 MB) [Process] => SwitchManager.cxx@513 : L5: Size of switch stats
- array = 2
SDK 15:45:07 C P F 1892 (20.52 MB 4.000000K) [Process Alert Work Element] => SDK@000I:
- MGK_AlertManager: creating thread for metric 0x00EC8350
SDK 15:45:07 C P T 1892 (20.52 MB) swrx_thread::create_thread => Creating new thread. (18 active and
- 0 blocked)
SDK 15:45:07 C P T 1892 (20.53 MB 12.000000K) swrx_thread::create_thread => Created thread 3140
SDK 15:45:07 C P T 3140 (20.59 MB 64.000000K) ThreadRun => SANITY: Begin internal thread wrapper
SDK 15:45:07 C P T 3140 (20.60 MB 8.000000K) ThreadRun => Start Thread "Alert Thread, Checking
- EGN.stats.connectivityport.numberlinkresets", Act: 18, Blkd: 0 (3140)
EGN 15:45:08 C P T 5404 (20.60 MB) [Process] => SwitchManager.cxx@730 : L5: ESN API returned 0
- handles for ip <192.168.101.40>
SDK 15:45:08 C P D 5404 (20.61 MB 8.000000K) [Checking Alert Value] => Checking Alert Metric =
- EGN.Alarms.Agent.StatisticsCollectionFailed, Keys = 100008008841ec00 Value = 1
SDK 15:45:08 C P D 5404 (20.61 MB) [Checking Alert Value] => Checking 5 thresholds for metric
- EGN.Alarms.Agent.StatisticsCollectionFailed"
No Events found!

Top