Enterprise Manager cannot reach TIM

[Manish Parikh]

I am seeing multiple ERRORs like these when I login to CEM but don't see any impact other than slow response time. TIM is a physical server and so is the MoM & WebView servers.

 

Enterprise Manager cannot reach TIM: http://xx.xx.xx.xx:80/wily/cem/tim/tess/settimtime?servertime=1464582240: 502 Proxy Error

 

Enterprise Manager cannot reach TIM: Error retrieving response http://xx.xx.xx.xx:80/wily/cem/tim/tess/settimtime?servertime=1464321240: Read timed out

Enterprise Manager cannot reach TIM: http://xx.xx.xx.xx:80/wily/cem/tim/tess/settimtime?servertime=1464150240: 502 Proxy Error

 

 

Thanks in advance for your help

[Hallett German]

Hi Manish:

This may be a temporary error due to a variety of issues. See 502 Bad Gateway Error (What It Is and How To Fix It)

And http://www.ca.com/us/support/ca-support-online/product-content/knowledgebase-articles/tec1602106.aspx

 

Make the suggested changes and report back if the thread can be closed or if there are additional questions

Thanks

Hal

[Gnanawathy Hareendran]

Hi Manish

A similar problem was reported by a different customer, which was related to "problem connecting to AD server. Network routing problem"

Thanks

Yanna

[Hallett German]

At this point, two answers have been provided and we are awaiting word on next steps...

[Manish Parikh]

Hallett_German

I have made the changes on non-prod TIMs, I will need to wait till after hours to work on the prod TIMs.

[Hallett German]

Hi Manish:

   Please  provide an update on this thread and if additional assistance is needed or can be marked as closed.

 

Thanks Hal German

[Hallett German]

Hi Manish:

    Did it make a difference on the non-prod TIMs?

Thanks

Hal German

[Lynn Williams]

Hi Manish,

The MOM log should have an ERROR for "Communication Error 502: 502 Proxy Error".

If that ERROR also includes "DNS lookup failure for: localhost" the likely root cause is that TIM itself is using DNS and cannot resolve the "localhost" correctly. If you get that fixed up it should then be OK

 

Regards,

 

Lynn

[Manish Parikh]

Lynn_Williams - not seeing any such ERROR (or any type of ERRORs for that matter) at all in MoM logs.

[Lynn Williams]

Try to enable or sync the TIM (if already enabled) and then check the log

If not in MOM log ERROR may be in EM log for the Collector running Tim Collection Service

[Hallett German]

Hi Manish:

Suggestions were provided and this thread has been quiet for two days. Marking as Answered. Please open a case if there are followup issues.

Thanks

Hal

[Manish Parikh]

Hallett_German Sorry for not posting on this thread yesterday. The issue still persists so I guess I'll open up a case. I am still getting ERRORs on CEM GUI

 

[Lynn Williams]

Hi Manish,

Is this a standard TIM or an MTP TIM?

Has this problem always existed since original TIM install?

Thanks

Lynn

[Manish Parikh]

Lynn,

This is standard TIM. I can tell you that this issue didn't exist when we initially installed TIM 9.6 at the beginning of the project ( a year ago). Then in late March/early June, I upgraded to 10.1. Now, I don't really remember if the problems started then (but I want to say no). Could this be an issue generated by applying openSSL and openSSH patches to the OS (where the TIM resides)? 

There is CA document that says ‘"If SSL communication is enabled between the TIM and EM/TESS, then do the additional step – “Configure the APM Enterprise Manager to use TLS for communicating to the TIM software"

 

Could this be contributing to the issue?

 

Thanks

Manish

[Lynn Williams]

Hi Manish,

Only if the unsecure port 80 has also been disabled on the TIM web server would it cause a problem & need the EM->TIM SSL comms enabled with "timTessCommunication.useSsl=1" in tess-default.properties and change the Monitor definttion for TIM to use port 443.

Check /etc/http/conf/httpd.conf for "Listen 80" uncommented.

Also this command should show httpd process listening on port 80: netstat -nap | grep 80 | grep LISTEN

 

Is it possible selinux has been enabled/enforced since the install?  If yes that could be impacting the web server.

Run getenforce to check.

 

Check the httpd logs in /etc/hppd/logs i.e. access_log & error_log.

If no errors there then it would seem EM request not getting to the TIM

If there are errors maybe also need to check the /opt/CA/APM/apmpacket/logs/apmpacketlog.txt & /opt/CA/APM/tim/logs/timlog.txt

(apmpacket was part of new performance improvement design in 9.7, so as you upgraded from 9.6 it could be relevant.)

 

If problem drags on we may need to have more detailed research via a support case.

 

Regards,

 

Lynn