โ09-26-2017 01:55 PM
Is anyone running a SnapLogic groundplex server running Windows hosted on AWS? On my orgโs grounplex server hosted on AWS, from time to time, I get a โconnection resetโ error on a File Reader snap when trying to read from any source (even SLDB) when the pipeline runs on the groundplex. Other times, the pipeline runs fine and the snap doesnโt error out. Iโve tried using โtcp traceโ to figure out why the connection reset happens, but I am not able to pinpoint the issue. Note that the error occurs very quickly after I execute the pipeline (within 1-2 seconds), so itโs not an issue of something timing out somewhere. Does anyone have any ideas about how to rectify this problem? Thanks in advance for any assistance.
โ04-26-2018 04:29 PM
Iโm experiencing the same behavior, but when running Linux hosted on AWS. Did you ever determine the root cause? Any help would be appreciated.
โ04-27-2018 05:22 AM
I have not received any responses, sorry. The issue still persists as well.
โ04-27-2018 04:47 PM
Disabling HTTP keepalive will help when the Groundplex to control plane connection is not reliable. Add this setting in global.properties and restart JCC process
jcc.jvm_options = -Dhttp.keepAlive=False -Dcom.ning.http.client.AsyncHttpClientConfig.allowPoolingConnection=false -Dcom.ning.http.client.AsyncHttpClientConfig.allowSslConnectionPool=false
If using the slpropz configuration file, add the above setting in Snaplex Global properties section in the Manager, that will initiate a rolling restart of the JCC nodes with the configuration update.
โ05-01-2018 05:18 AM
Thanks for the suggestion. I have added this setting to my global.properties file and I have not noticed the error since it has been in place (~ 24 hours).
Fingers crossed that it will be a permanent fix.
Thanks again!