Some jobs Abort right after the Startup of the Manager for SAP with the following kind of errors in universe.log:
|ERROR|X|SAP|pid=p.t| u_RfcUTF8ToSAPUC | Error [22] while trying to convert [330]
|ERROR|X|SAP|pid=p.t| uxsap_get_connect_param | Error while trying to convert CLIENT value [330]
|ERROR|X|SAP|pid=p.t| uxsap_open_comm | SAP system [*/*/*]: uxsap_get_connect_param fails returned error -1
|ERROR|X|SAP|pid=p.t| uxsap_connect | uxsap_open_comm error: unable to open the RFC handle
|INFO |X|SAP|pid=p.t| uxsap_connect | RFC connection fails for SAP system [*/*/*]
|INFO |X|SAP|pid=p.t| uxsap_connect_with_retry | Trying a new connection in 3 seconds... Attempt 2/6
|ERROR|X|SAP|pid=p.t2| u_RfcSAPUCToUTF8 | Error [22] while trying to convert a string
|ERROR|X|SAP|pid=p.t3| u_RfcSAPUCToUTF8 | Error [22] while trying to convert a string
|INFO |X|SAP|pid=p.t4| uxsap_find_jobs_alljnl | No job found in memory for key [1/COMPANY/X/NODE/6654808/AR3]
|INFO |X|SAP|pid=p.t5| uxsap_find_jobs_alljnl | No job found in memory for key [1/COMPANY/X/NODE/6654812/AR3]
In job log, we could find a trace like this:
===================== SUBMISSION ERROR ===================== unable to open job [UPROCNAME] ------------------------------------------------------------ BAPI error number : id : message : ============================================================
Relaunching them allows to submit them correctly.
Release : 6.x
Component : Manager for SAP
Defect: due to an error within the libnwrfc that seems to take a few seconds to correctly load.
To avoid this behaviour, a small retry mechanism, based on UXSAP_NB_CONNECT_RETRY variable is taken into account.
Relaunch the aborted Jobs or delay its startup before starting the Node or Manager for SAP.
Update to a fix version listed below or a newer version if available.
Fix version(s):
Component: Manager.for.SAP
Dollar Universe 6.10.101 - Planned Release First half July 2022