Details
-
Type:
Bug
-
Status:
Verified
-
Priority:
Major
-
Resolution: Unresolved
-
Affects Version/s: 6.1.20 EE GA2, 6.1.X EE
-
Fix Version/s: None
-
Component/s: Upgrades
-
Environment:Sharding
scheduler.enabled=true
-
Fix Priority:4
-
Similar Issues:
Description
This is only reproducible if you try an upgrade from 6.0.11 (or below) to 6.0.12 (or above)
Steps to reproduce it:
1- Startup Liferay portal 6.0 SP1 using sharding
2- Create at least 2 virtual host
3- Upgrade to 6.1GA2 using the changes of the issue LPS-31817 (it didn't caused the problem, it just brought it to light)
The following exception is thrown:
16:04:29,314 INFO [pool-2-thread-1][UpgradeProcess:175] Upgrading com.liferay.portal.upgrade.v6_0_12.UpgradeScheduler
16:04:29,406 ERROR [pool-2-thread-1][MainServlet:217] com.liferay.portal.kernel.events.ActionException: java.sql.SQLSyntaxErrorException: ORA-00904: "STATE_": invalid identifier
com.liferay.portal.kernel.events.ActionException: java.sql.SQLSyntaxErrorException: ORA-00904: "STATE_": invalid identifier
The problem is that the portal is trying to upgrade the Quartz tables in all the shards but it's only necessary in the shard by default.
Workaround:
Execute the following query in all the shards except for the shard by default before performing the upgrade:
CREATE TABLE quartz_job_details (
JOB_NAME varchar(200) NOT NULL,
JOB_GROUP varchar(200) NOT NULL,
JOB_DATA longblob
)
(Mysql syntax, adapt it to yours)