Someone set up a server at my workstation, which allows our Postgres servers to retrieve data from our MS SQL servers using an external data wrapper, with something like this:
https://github.com/GeoffMontee/tds_fdw
It works pretty cool until we try to pull out or request large volumes of records in 10,000 seconds. Postgres just kicks us like this:
apitransform=
NOTICE: DB-Library notice: Msg #: 5701, Msg state: 2, Msg: Changed database context to 'master'., Server: abcdefg, Process: , Line: 1, Level: 0
NOTICE: DB-Library notice: Msg #: 5703, Msg state: 1, Msg: Changed language setting to us_english., Server: abcdefg, Process: , Line: 1, Level: 0
NOTICE: DB-Library notice: Msg #: 5701, Msg state: 1, Msg: Changed database context to 'abcdefg_staging'., Server: abcdefg, Process: , Line: 1, Level: 0
NOTICE: DB-Library notice: Msg #: 5701, Msg state: 2, Msg: Changed database context to 'master'., Server: abcdefg, Process: , Line: 1, Level: 0
NOTICE: DB-Library notice: Msg #: 5703, Msg state: 1, Msg: Changed language setting to us_english., Server: abcdefg, Process: , Line: 1, Level: 0
NOTICE: DB-Library notice: Msg #: 5701, Msg state: 1, Msg: Changed database context to 'abcdefg_staging'., Server: abcdefg, Process: , Line: 1, Level: 0
The connection to the server was lost. Attempting reset: Succeeded.
We are having trouble filling out our records because it seems like one of these hard limits exists.
One key is that when FreeTDS, when set to 4.2, we can query approximately fine, or at least be able to count records. However, we ran into UTF character encoding problems. So, we set it to 7.0 and 8.0, and this solves the UTF problem, but now there is a limit on how much we can pull at a time, and the limit is unpredictable and often breaks.
Can anyone know what a deal is? Is it somewhere in a small setting?