vRSLCM calls api with stale data when upgrade is retried through VRSLCM and hence VRNI Upgrade fails in pre-checks
Error message seen:
com.vmware.vrealize.lcm.plugin.core.vrni.common.exception.VRNIUpgradeCheckStatusException: Error occurred while checking upgrade pre-check status with IP - Pre-check message :
{ "msg" : "HBase or HDFS is unhealthy", "type" : "INFO", "status" : "FAIL", "id" : "HBaseHdfsHealthCheckTask", "title" : "HBase and HDFS Health Check", "consentMsg" : null}Run below steps on platform1/single node 1) Find pre-checks UUID, version number and upgrade type using below command: sudo grep "Start upgrade checks for id" /var/log/arkin/pre_checks.log | tail -1 | awk '{print $6 $8 $10}' | sed 's/,/ /g' Example Output: bdddd00a-XXXX-XXXX-XXXX-XXXXXXXXXXXX 6.6.0.1648892975 OFFLINE 2) Trigger pre-checks using the output above in the below command: sh /home/ubuntu/build-target/pre-upgrade/pre.sh <Output of 1st command> Example Command: sh /home/ubuntu/build-target/pre-upgrade/pre.sh bdddd00a-XXXX-XXXX-XXXX-XXXXXXXXXXXX 6.6.0.1648892975 OFFLINE 3) Verify if pre-checks passed Command: sudo tail ~/logs/pre_checks.log Output: last few lines of file(~/logs/pre_checks.log) should contain such lines: Checking if HDFS is healthy Checking if HBase is healthy Done upgrade checks for id bdddd00a-XXXX-XXXX-XXXX-XXXXXXXXXXXX, version 6.6.0.1648892975, upgradeType OFFLINE
Setups which use vRLCM 8.7.0 and 8.8.0 and upgrade fails during reattempt due to pre-checks failure.