# Flowable Engine Upgrade
**URL:** https://forum.flowable.org/t/flowable-engine-upgrade/6415
**Category:** Flowable Engine
**Created:** [July 22, 2020, 2:20am UTC](https://forum.flowable.org/t/flowable-engine-upgrade/6415 "2020-07-22T02:20:09Z")
**Posts on this page:** 4
**Page:** 1
### Author:  [@aytay1](https://forum.flowable.org/u/aytay1)
#### Post date: [July 22, 2020, 2:20am UTC](https://forum.flowable.org/t/flowable-engine-upgrade/6415/1 "2020-07-22T02:20:09Z")
Our application was running on flowable version 6.4.2 and we decided to upgrade to flowable version 6.5.0. We are using Java 8 with Spring Boot framework v2.2.1.RELEASE. Our database is Oracle 12c. After I upgraded the flowable version in the pom.xml file from 6.4.2 to 6.5.0, and ran the upgrade script from this site ([https://github.com/flowable/flowable-engine/blob/master/distro/sql/upgrade/all/flowable.oracle.upgradestep.6.4.2.to.6.5.0.all.sql](https://github.com/flowable/flowable-engine/blob/master/distro/sql/upgrade/all/flowable.oracle.upgradestep.6.4.2.to.6.5.0.all.sql)), I deployed the flowable web to Oracle Weblogic 12c.
If I start a new process, the flowable works fine. However, for all existing ongoing processes are not working at all. When I try to transit the flow, it throws me this exception **deployment ‘476749f6-c1ee-11ea-9311-000c29c5dcb3’ didn’t put process definition ‘47d63869-c1ee-11ea-9311-000c29c5dcb3’ in the cache**. One workaround was that I removed all existing data (ACT\_RU\_TASK, etc) and recreate the task. However it’s not ideal since we might want to upgrade the flowable version on production environment, and we can’t afford to remove existing data. May I know if there is any alternative solution for this?
Thank you.
---
### Author:  [@joram](https://forum.flowable.org/u/joram)
#### Post date: [July 30, 2020, 4:47pm UTC](https://forum.flowable.org/t/flowable-engine-upgrade/6415/2 "2020-07-30T16:47:31Z")
This exception can happen only when the reading the definition and deploying it in memory didn’t put it in the cache. Which normally shouldn’t happen, unless there’s maybe massive contention on the cache. Are you using any custom or configured definition cache? Cause by default, the cache is unlimited in size.
Also: did you reboot the server after the upgrade?
---
### Author:  [@aytay1](https://forum.flowable.org/u/aytay1)
#### Post date: [September 2, 2020, 9:58am UTC](https://forum.flowable.org/t/flowable-engine-upgrade/6415/3 "2020-09-02T09:58:38Z")
We are not using any custom / configured definition cache. We never put any property related to flowable in the properties file.
We reboot the server after the upgrade.
---
### Author:  [@joram](https://forum.flowable.org/u/joram)
#### Post date: [September 7, 2020, 11:33am UTC](https://forum.flowable.org/t/flowable-engine-upgrade/6415/4 "2020-09-07T11:33:08Z")
Looking again through the code, I can’t see how it can happen in the default setup … can you maybe share logs before and after when that line from above is happening? Preferably on debug level, with the mybatis sql shown.