Analytics Things stopped connecting after IP address of the VM changed

[SajidPatel]

I have Thingworx platform and Thingworx Analytics installed in a CentOS VM on my laptop. It was working fine until, the IP address of the VM changed (or so I suspect, may be something else also might have affected this).

Please see the attached snapshots - all configurations are with localhost. I haven't used any IP addresses while configuring the Analytics and everything is on the same VM. Is there a way to get edge logs to see whats the error? I do not see anything useful in the Thingworx logs or Analytics logs.

[cmorfin]

Hi @SajidPatel 

From the properties file it appears that ThingWorx Analytics connect to ThingWorx using the hostname centos7.

Given that this is a VM I am guessing this is a local name resolution via /etc/hosts file.

So maybe the ip settings in /etc/hosts for this hostname is not up to date anymore.

You can try ping centos7 in a terminal to check if it responds.

Regarding log files, those depends on the version of ThingWorx Analytics used, the article https://www.ptc.com/en/support/article?n=CS268761 will tell you where to find them based on the version.

Hope this helps

Kind regards

Christophe

[SajidPatel]

@cmorfin Thanks for responding

 centos7 is the localhost. ping works fine, please see the attached screenshot.

[cmorfin]

Hi @SajidPatel 

Can you specify the version of ThingWorx Analytics you are using and can you upload all the logs from <ThingWorxAnalytics>/data/logs folder ?

Thanks

Christophe

[SajidPatel]

Please find the logs attached.

I have ThingWorx 8.3.4-b993

& ThingWorx Analytics Manager Framework 8.3.3 b1036 

& Thingworx Analytics Server 8.3.3 (Linux - CentOS)

[cmorfin]

Hi @SajidPatel 

Thank you for this.

The logs do not shed any light so far.

Could you please do the following:

- execute <ThingWorxAnalytics>/bin/twas.sh stop

- confirm that no ThingWorx Analytics Server processes are running with ps -ef | grep java

if some orphan processes remain, simply kill them

- delete the content of <ThingWorxAnalytics>/data/logs

- restart the server:  <ThingWorxAnalytics>/bin/twas.sh start

- wait 5 min

- if the AnalyticsGateway thing is not connected in Composer after this time, do the following

- execute  <ThingWorxAnalytics>/bin/twas.sh status > /tmp/twaStatus.txt

Send me :

- the /tmp/twaStatus.txt

- a zip of <ThingWorxAnalytics>/data/logs

- a zip of <ThingWorxAnalytics>/config

Thank you 

Kind regards

Christophe

[SajidPatel]

@cmorfin 

Please find attached logs.zip - it includes all that you asked for + zookeeper.out. There were no log files generated in data log. Also included the yml files where I attempted to bump up the logging verbosity.

The twas status was this:

[root@centos7 bin]# ./twas.sh status
ThingWorx Analytics Server (TWAS) Service is not running ...

[SajidPatel · Accepted solution]

Thanks for your time @cmorfin 

Due to these investigations, I was able to find the root cause of the issue. The Java JRE path was hardcoded with specific Java JRE version. It was updated when I installed some other component on the VM as a result breaking the services. I have updated the services and it is working fine now.

Thanks! 

[slangley]

Hi @SajidPatel.

Since you found the solution, please mark your last post as the Accepted Solution for the benefit of others with the same issue.

Regards.

--Sharon