Source code retrieve in DataGrip having problems

I have tried really hard to like DataGrip several times. It seems to be getting better, but there is one huge non-starter for me that doesn't seem to be getting fixed. I can not seem to get to the source code of package, views and other objects. It appears that a synchronization must complete successfully in order for this to happen? I work with a large Oracle 11G database. Not just a lot of data, but a lot of objects in one schema. I have set my JVM settings to -Xmx1024m and it still will not complete the sync after several hours.

I assume this is because you are syncing all of the objects with their source? While I understand the idea for autocomplete, I still need the ability to view source on an object even if the object is not synced. Plus if you are only showing me the synced object, what happens if the object source changes? Does a re-sync have to happen before seeing the real source? This re-sync process for me seems to take hours and while happening in the background, it still appears to keep me from seeing the current source. If this is not the case, what could I be doing incorrectly? Could the "show source" not go to the database directly to get the current source while using synced data for autocomplete?

For me, this the the last product that I have resisted switching. I use IntelliJ, Rider, PyCharm, and WebStorm daily and would not give any of them up for any other tool. PLEASE help me like DataGrip as well!!!

Database: Oracle 11G

OS: Mac OS Mojave and Ubuntu 18.04 both

Datagrip version: 2018.3



Hi! This is obviously not OK.

Please send us Logs and SQL logs to

They can be found in Help menu (Show logs in...)


Hi Michael, you can make the syncing process faster by granting the "SELECT ANY DICTIONARY" permission  to the account you're logging in DataGrip with (this permission allows to use the system catalog that is much faster than the user's one).



I have submitted the logs. Did you receive? 


I will try that...Thanks


Hello Michael,

the first problem I've found in the log that the JVM process has small heap size. Moreover, when the introspection performs, there are two JVM processes are involved: the main process and the JDBC process; both of them should have large heap segments for introspecting so large database.

I can explain how to configure the memory limitation, but the explanation depends on how do you start DataGrip: using ToolBox or not?


I do install Datagrip using the toolbox. I generally do not launch using the toolbox though. I launch using the Mac launcher or .desktop file in Ubuntu. Is there a different configuration that should be used beyond adding -Xmx1024m into the connection VM Options?


If you launch DG not from Toolbox, just use menu Help / Edit Custom VM options to open the JVM settings file, and add/edit the Xmx option in order to set 4096 mb for the main DataGrip (really the Intellij platform) process.

In order to configure the JDBC process, open your data source configuration, then switch to the Advanced tab, then find the JVM options line at the bottom of the form and add -Xmx4096m here.

Then try to introspect your Oracle database again.

If it failed, please provide me with the SQL log (menu Help / Show SQL log in…) and the normal log files (stack traces in subdirectories are not needed). 


Please sign in to leave a comment.