Slow process to clone with source control

jescan

Hello, I installed Source Control for Oracle and I am working with oracle database 11 xe on a azure vm with ubuntu 16, but I see that the process is very slow when I try to clone the github repository with my database, is this delay normal?

Alex B

Hi @jescan,

The performance in this scenario is going to be impacted by several things including the connection to both the Azure VM where the database resides and to where your source control system is located and the number of objects that are in the schema(s) you have added to the project.

Source Control for Oracle works in a similar way to SQL Source Control as described on this page https://documentation.red-gate.com/soc7/reference-information/how-sql-source-control-works-behind-the-scenes and so there are various actions going on when you set up the project including reading information from the database itself and also connecting to the repository.  

If there are a large number of objects this also impacts the speed as the comparison process performance is largely related to the number of  objects.

Kind regards,
Alex

jescan

Hello @Alex B

Thank you for your answer, the topic is clearer, but I tell you, that for these tests it is a database with two tables of 4 fields each and 5 records. It is true, the VM is on Azure but with this load it should not take long.

I also tell you that the source control is installed on a local PC and what I am trying to do is deploy the database in the azure VM from the Github repository.

On the other hand, I also told you that I have worked with source control for sql server and it has been luxurious, without any problem. That is why I am surprised that this happens with oracle, it is more now I additionally get Error ORA-010301: insufficient privileges

regards
Jesus

Alex B

Hi @jescan,

The tool uses various data dictionaries to get information it uses.  It will try to use the dba_* version first, but if it doesn't have permission to that, then it will fall back and use the all_* version which, in addition to having to wait for the first to fail, also takes longer.  It sounds like this may be what is happening.

The data dictionaries we use are listed on the bottom of this page https://documentation.red-gate.com/sco5/requirements/permissions-required-to-use-schema-compare-for-oracle  and it also goes over the other permissions requirements.

Kind regards,
Alex