Hi all - I know this might sound like a problem that had occured at least 50 times before, but I've read just about ALL threads on this subject, to no avail.
We have just recently migrated our transformations and jobs to a database repository; and unfortunately we have been getting timeouts to the repository ever since. That is, we can connect to the repo and start working in Spoon just fine - for about 10 minutes - but after this the connection to the repo drops and so if we want to load or save, Spoon locks up, throwing this error in the end:
We can only continue work by Tools --> Disconnecting and Tools --> Connecting back to the repo again.
The greater problem is that the problem also persists in Kitchen. When running a longer job - say longer than this 5-10 minutes mentioned earlier - Kitchen fails to fetch the oncoming transformation and drops an error. I will post the error log at the end of this post. It seems however that only the repository connection times out. Data can still be output to the database as long as the current transformation is running.
I have been going through just about ALL threads on these forums and even stackoverflow.
Most of them end with this link: http://wiki.pentaho.com/display/EAI/MySQL
However, I have already set net_write_timeout (and net_read_timeout) to 1800 and can confirm that the variable is set to stay even after a server restart (we have also created a my.cnf specifically for this).
I have even tried updating the mysql connector from 5.1.25 to 5.1.27 - unfortunately, this hasn't changed anything either.
And please see the most recent error log attached. All previous ones look exactly the same.
Does anyone have an idea why the repo times out and what I can do?
errorlog.txt
We have just recently migrated our transformations and jobs to a database repository; and unfortunately we have been getting timeouts to the repository ever since. That is, we can connect to the repo and start working in Spoon just fine - for about 10 minutes - but after this the connection to the repo drops and so if we want to load or save, Spoon locks up, throwing this error in the end:
Code:
An error occured loading the directory tree from the repository
ERROR executing query in part[OQ exec query]
Communications link failure
The last packet successfully received from the server was 635 440 milliseconds ago. The last packet sent successfully to the server was 21 301 milliseconds ago.The greater problem is that the problem also persists in Kitchen. When running a longer job - say longer than this 5-10 minutes mentioned earlier - Kitchen fails to fetch the oncoming transformation and drops an error. I will post the error log at the end of this post. It seems however that only the repository connection times out. Data can still be output to the database as long as the current transformation is running.
I have been going through just about ALL threads on these forums and even stackoverflow.
Most of them end with this link: http://wiki.pentaho.com/display/EAI/MySQL
However, I have already set net_write_timeout (and net_read_timeout) to 1800 and can confirm that the variable is set to stay even after a server restart (we have also created a my.cnf specifically for this).
I have even tried updating the mysql connector from 5.1.25 to 5.1.27 - unfortunately, this hasn't changed anything either.
And please see the most recent error log attached. All previous ones look exactly the same.
Does anyone have an idea why the repo times out and what I can do?
errorlog.txt