cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 

This product reached the end of support date on March 31, 2021.

Agent-> Collector Timeout ssl handshake

jan-hendrik_wol
Organizer

Hi,

is it possible to reduce the timeout of the SSL Handshake between Agent and Collector?

We had a big issue in our network environment yesterday and most of our systems needed a reboot.
Some of our JVMs running Dynatrace had problems to come up again.

The reason is, that our wrapper restartet the application before Dynatrace could make its handshake.

INFO  
| jvm 4 | 2017/01/11 14:48:05 | 2017-01-11 14:48:05
[d7b5681a] info [native] Detected application server: Apache
Tomcat


INFO
| jvm 4 | 2017/01/11 14:48:05 | 2017-01-11 14:48:05
[d7b5681a] info [native] dynaTrace Bootstrap Agent 6.3.0.1305


INFO
| jvm 4 | 2017/01/11 14:48:05 | 2017-01-11 14:48:05
[d7b5681a] info [native] Using /usr/local/tomcat/dynatrace as
Agent storage


INFO
| jvm 4 | 2017/01/11 14:48:05 | 2017-01-11 14:48:05
[d7b5681a] info [native] Log file is
/usr/local/tomcat/dynatrace/log/dt_mad-p3_bootstrap_10418.0.log


INFO
| jvm 4 | 2017/01/11 14:48:05 | 2017-01-11 14:48:05
[d7b5681a] info [native] Trying to connect to Collector for
up to 20 seconds


INFO
| jvm 4 | 2017/01/11 14:48:05 | 2017-01-11 14:48:05
[d7b5681a] info [native] Connected to Collector
10.196.52.9:9998


ERROR
| wrapper | 2017/01/11 14:48:34 | Startup failed: Timed out waiting for a
signal from the JVM.


ERROR
| wrapper | 2017/01/11 14:48:35 | JVM did not exit on request,
termination requested.


STATUS
| wrapper | 2017/01/11 14:48:35 | JVM received a signal SIGKILL (9).


STATUS
| wrapper | 2017/01/11 14:48:35 | JVM process is gone.


STATUS
| wrapper | 2017/01/11 14:48:35 | JVM exited after being requested to
terminate.


To be honest, the handshake was working again, after restarting the collector, but we lost some time of investigating, while our systems were standing still.

I found this thread:
https://answers.dynatrace.com/questions/105497/age...

But this variable seems to work just for the initial connection and not the handshake.

Regards,

Jan

1 REPLY 1

c_schwarzbauer
Dynatrace Champion
Dynatrace Champion

hi Jan,

I honestly don't see a connection to any network issue or handshake problem (also I'm pretty sure that it's not SSL at all) from the above information.

@ 14:48:05 the agent says it's trying to connect to the collector and in the same second it says that it's connected, so I don't see any issue around the handshake here yet.

what I would assume more likely is that due to your outage all the JVMs had to be restarted and that the Collector was then too busy with instrumenting all those JVMs that the startup was delayed beyond what your wrapper allows (I assume 30 seconds?). does that sound reasonable?

if that would be the case I have to ask: how many agents do you have on one collector? it might help here to use more collectors and balance the agents between them. or you could possibly also adapt the wrapper timeout.

HTH, Christian