zos APM Bridge does not connect to KAFKA on two of the four LPARS in the PLEX
Article ID: 389273
Updated On:
Products
Issue/Introduction
The ZABPROC task does not successfully connect to KAFKA on two of the 4 LPARs on the same sysplex. All 4 LPARs are attempting to connect to the same KAFKA.
DEBUG c.b.m.z.console.OpsConsolePublisher - ZAB heartbeat issued.
INFO com.broadcom.mfd.zab.ZABApplication - Log directory location: /produc
INFO com.broadcom.mfd.zab.ZABApplication - Started ZABApplication in 5.034
After numerous attempts to connect, the Metrics scheduler closes:
INFO o.a.kafka.common.metrics.Metrics - Metrics scheduler closed
INFO o.a.kafka.common.metrics.Metrics - Closing reporter org.apache.kaf
INFO o.a.kafka.common.metrics.Metrics - Metrics reporters closed
INFO o.a.kafka.common.utils.AppInfoParser - App info kafka.producer for pro
WARN c.b.mfd.zab.sink.KafkaSinkService - Attempt 1: Exception while send
On the PLEXs that successfully connect to Kafka, the following message appears in the ZABPROC log.
BPXM023I (ZABUSER) ZABM002I z/OS APM Bridge is connected to CCS Data Mover and APM Kafka
This message never appears on the unsuccessful LPARS.
Cause
Networking Issue
Resolution
The first thing to check is the network. Recommend issuing a PING from each of the failing LPARs to the KAFKA server.
The PING was successful on the first two servers, as expected, but did not return responses on the two where the connection could not be made. The Network team had to be involved to fix a problem caused by a recent network hardware upgrade, and once they did, the connection established successfully.
If connectivity cannot be established, and it is determined that there is no network problem via PING or TRACERT, then open a Support case and provide the ZABPROC log for investigation.
Feedback
