DA Upgrade Fails Requiring 126GB Disk Space With Only 71GB Available

Products

• Network Observability
• CA Performance Management

Issue/Introduction

During a Data Aggregator upgrade, the process fails and reports insufficient disk space, requesting significantly more storage than available for backing up existing data․​​​‌​‍

ERROR MESSAGE: DA upgrade fails, says it needs 126GB of disk space․

SYMPTOMS:

• Upgrade from 24․3․x halts

• Pre-flight disk space check fails

• Backup requirement exceeds available free space

CONTEXT: Occurs during the Data Aggregator upgrade process․

Environment

DX NetOps PM: Any version

Cause

A very large java_pidxxxxxx․hprof heap dump file exists in the apache-karaf directory․ The upgrade process attempts to back up this file along with existing data, artificially inflating the required disk space․

Resolution

PREREQUISITES:

• Command line access to the Data Aggregator server

• Sufficient permissions to remove files

STEPS:

1․ LOCATE LARGE HEAP DUMP FILES

Path: /opt/IMDataAggregator/apache-karaf-*/ Command: ls -lh java_pid*․hprof

EXPECTED: Identifies large Java heap dump files consuming excessive space․

2․ REMOVE UNNECESSARY HEAP DUMP FILES

Command: rm /opt/IMDataAggregator/apache-karaf-*/java_pidxxxxxx․hprof

EXPECTED: File is removed, freeing up sufficient disk space․ NOTE: Replace xxxxxx with the actual process ID found in the file name․

3․ RETRY THE UPGRADE

Restart the Data Aggregator upgrade process․

EXPECTED: Upgrade completes successfully without disk space errors․

VERIFY SUCCESS:

• Disk space requirements are met

• NetOps upgrade finishes successfully

Additional Information

Find largest directories:

du -h --max-depth=1 /path/to/start/from | sort -rh | head -n 10

du -h --max-depth=1 /opt/IMDataAggregator/ | sort -rh | head -n 10

 

Find largest files in a directory:

find /path/to/start/from -type f -printf '%s %p\n'| sort -nr | head -25

find /opt/IMDataAggregator/ -type f -printf '%s %p\n'| sort -nr | head -25