We have CloverETL Server 3.4.0.23M1 running on an instance in EC2, backed by EBS.
Today the instance became unresponsive and we had to restart it. On restarting it, the server came back up like a fresh install. No sandboxes, no schedules, and no execution history.
The job files are all still on the server and we were able to load them back into CloverETL Server. We also manually recreated the schedules.
Why did this happen? Where does CloverETL store the things we lost (sandbox links, schedules, and execution history)?
Hi nct2013,
At first, I highly recommend not to use milestone version for production purposes. First milestone is the most dangerous one because is far from production state. Typically in M1 we are releasing brand new functions as preview for customers. Also, 3.4M1 was released 2012-12-19 so it is really outdated version.
Regarding lost configuration. Can you please specify your DB configuration? I would guess you use Derby, which is simple DB engine not recommended for production - http://doc.cloveretl.com/documentation/ … ation.html Derby stores data into single file, see http://doc.cloveretl.com/documentation/ … rence.html for Derby configuration. This file was probably broken during server crash. If it is the case, I am afraid it will be hard to recover.
I hope this helps.
Our database URL on the CloverETL config page includes “Derby”, so yes, I believe that’s what we are using.
I will upgrade our version of CloverETL and report back if we still see this issue.