Various properties files and ant scripts involved in upgrade have default timeouts of 6 minutes. For large/slow systems, Admin may not respond within this time limt, so TCT will abort the upgrade.
1. Revert the system to your pre-upgrade backup. 2. Re-install BPM 4.3 (if not installed before taking backup) 3. Using any standard text editor, increase the value of httpConnectionTimeout settings in the following files (all under TIBCO_HOME):
Note: on Linux systems, it may be easier to patch the files using sed:
# increase timeouts from 6 to 60 minutes (ant scripts) cd TIBCO_HOME for x in $(grep -Rl httpConnectionTimeout bpm/4.3/scripts/*|grep -E 'xml$') ; do sed -i.orig -re 's/(httpConnectionTimeout" type="string" value)="360000"/\1="3600000"/g' $x && echo "Patched $x" done
# increase timeouts from 6 to 60 minutes (remote_props) for x in {administrator/3.4/scripts/governance/util/reloadGOCache,bpm/4.3/scripts/deployer}/remote_props.properties ; do sed -i.orig -re 's/(httpConnectionTimeout)=360000/\1=3600000/' $x && echo "Patched $x" done
4. Perform the required upgrade steps using TCT (per documentation)
Issue/Introduction
Some tasks in BPM 4.3 upgrade scripts have a 6 minute timeout
This is too short for large/slow systems, preventing upgrade to 4.3