My UPA application connects to TREP but is getting occasional timeouts followed by channel reconnections. The logs from my API are below.
2019/03/05-14:30:41.884857 +0000 Status code <RSSL_SC_TIMEOUT> data state <RSSL_DATA_SUSPECT> stream state <RSSL_STREAM_OPEN> for </ELEKTRON_DD/CFCB.OQ> stream id  <State: Open/Suspect/Timeout - text: "Request timed out."> 2019/03/05-14:32:27.953423 +0000: Connection reconnecting Error: (</local/jenkins/workspace/ESDKCore/OS/RH6-64/esdk/Cpp-C/Eta/Impl/Transport/rsslSocketTransportImpl.c:3168> Error: 1002 ipcWrite() failed. System errno: (104)>
The timeout may not be related to the channel reconnection but it's the only other unusual activity near the time of the reconnect.
I don't have access to the logs at the TREP installation yet.
What normally causes these kinds of errors ?
"Request timed out." could be a request timeout in the UPA Value-Added library (Watch List). The default value is 15 seconds. The requestTimeout option is defined in RsslConsumerWatchlistOptions.
If the UPA Value-Added library doesn't receive any response back from the infrastructure for 15 seconds after sending an item request, the UPA Value-Added library will send this status message to the application. Then, it will close and resubscribe to the same item. You may need to increase the requestTimeout value.
Regarding disconnection, the system error number is 104 which is Connection reset by peer.
#define ECONNRESET 104 /* Connection reset by peer */
We may need to verify the ADS log for the reason for disconnection.