Gilles Darold
Gilles Darold
I would like to help but this is a network issue, you may have a firewall or possibly at Azure side that is resetting the connection leading to this timeout....
Hi, So you are using 24 cores to export this table, what is your CPU and memory resources? Also what is the value of DATA_LIMIT or the -L option of...
Thanks, then you can give a try to options `-J 4 -j 3 -L 5000` to not overcommit resources. It also depends of your geometry size, looks like it is...
How about the export using `--ora2pg_speed`, does it show the same slow down? Your test last test show that after resources are freed it restart at the same speed.
Thanks a lot for all these information. The fetch rate decrease is strange because we are on the same table, usually DBD::Oracle switch to a per single fetch mode when...
yes if you can test several setting to see how things changes this will be a great help.
The settings about caching/prefetching is done in the call to prepare() at lines 14093 and 14121. In your case this will be at line 14093 as you have ORACLE_COPIES =...
You can also give a try to export the table without exporting the geometry column to see if the problem is related to this column. You can do that by...
Do you know what happen to memory during the export? If you can post the gzip file resulting from the following command executed before the export and stopped at end:...
With your test ora2pg forks 48 processes (6x8) this is probably too much for your 12 cores.