Update: The timeacquired and lastrun are stuck at 0 rather than get updated with a timestamp(2) value after initial run. If stuck at 0, no refresh occurs.
After altering timestamp, I had to update with a valid date. e.g. update ROLLER.ROLLER_TASKLOCK set timeacquired = '01-APR-11 08.22.06.040000 AM', lastrun='01-APR-11 08.22.06.040000 AM'; On Fri, Apr 1, 2011 at 1:29 AM, Van Rogers-Ho <[email protected]> wrote: > Ok. I was able to get it working. Looks like it's related to Oracle > specific date precision. > > [jira] Commented: (ROL-1760) Scheduled tasks do not run on Oracle (or other > DBs with high precision timestamps) > > I ran following for Oracle: > > update ROLLER.ROLLER_TASKLOCK > set timeacquired = null, > lastrun=null; > > commit; > > alter table roller_tasklock modify timeacquired timestamp(2) with time > zone; > alter table roller_tasklock modify lastrun timestamp(2) with time zone; > I'm getting recent entries on my front page and /planetrss now. > > Thanks, > Van > > On Thu, Mar 31, 2011 at 7:16 PM, Van Rogers-Ho <[email protected]>wrote: > >> Hi, >> >> I'm hoping someone can provide some files and logs to look at to diagnose >> why RefreshRollerPlanetTask:runTask is no longer running. >> It looks like a few weeks ago (mid march), it stopped. I looked at old >> roller.log and noted the following: >> >> >> INFO 2011-03-13 00:05:00,067 RefreshRollerPlanetTask:runTask - Refreshing >> Planet subscriptions >> >> In latest roller.log, I don't see this message, nor do I see any type of >> errors. Is there a way to manually fire off the task to check for errors in >> logs? Any planet specific log? >> >> I also cleared out my ROLLER.RAG_ENTRY table to get rid of any junk. >> >> Running: 4.0.1 on WLS 11g, and Oracle db. >> >> Thanks, >> Van >> > >
