i got a problem on our Netweaver Portal Server. In folder sapmnt\<SID>\<SID>##\j2ee\cluster\server0 there are a lot of dev_jrfc,trc files with 8 mb (on a single day some gigabytes..) and hdd space run short. Usually it should help to change -Djrfc.trace, -Djco.trace_level, -Djco.jarm to 0 in configtool and restart the cluster. But in my case it doesnt work, it takes no effect. whats wrong?
Into the trc logs are only written some RFCIO_ERROR_INTERNAL(3) but it doesnt matter because connections are working. So we can disable the trace. Here the instance_ID settings
From above document:
dev_jrfc.trc : This is the RFC error log that is always produced if an RFC error occurs. It cannot be switched off.