Comments by Evilfish2000

Evilfish2000 23-Nov-23 5:32am View

Maxim thank you for your suggestion. I will test this at some point. However, work suddenly took a turn so this project is on hold. Sorry for not informing of this until now.

Evilfish2000 24-Apr-23 10:47am View

It's true that the answer is not strictly correct with the answer. But, I was trying to read the headers of emails and the meme kit does exactly that.

Evilfish2000 24-Apr-23 6:45am View

Could you post that comment as a solution cause this is just what I needed and more.

Evilfish2000 22-Dec-22 2:53am View

We have been experimenting with batch limits between 500-2000. Currently it is set to 500, which have been our normal one. This was chosen initially, as it seems to be the one that performs best bandwidth wise. I've seen more comments about removing the index when doing the insert, which is a strategy I have in mind as the next step. We are still trying test with just rebuilds or reorganizations first, though our test environment seems to have gone on Christmas holiday already.

Evilfish2000 22-Dec-22 2:49am View

No, there is no truncation.

Evilfish2000 22-Dec-22 2:48am View

We only have one location with a cursor that is a manual maintenance task that is executed only when necessary. Otherwise, I avoid using such constructs.

Evilfish2000 20-Dec-22 8:11am View

All sites load data into several tables, but all sites have the same tables. So all site load data from their TableA to server TableA and so on. I am currently investigating that rebuilding index may help with the issue.

Evilfish2000 20-Dec-22 8:08am View

Unlikely. We have a 30 second timeout on the bulkcopy and we have seen processes up to half an hour complete without issue from sites half way around the world (Good suggestion though). I have however found a promising lead. One of my colleagues did a batch today again, which resulted in the same failure. We then looked into some server processes as another colleague suggestion that a cleanup routine might be why we can do it the next day. We did have a routine, that among other stuff rebuilds indexes and handles some logs. We ran that and retried again, where it completed all tables except one which we currently think was just a fluke that we ran again and it worked. So we are working on doing some automatic rebuilds on indexes upon table adds and see if that should help with the issue. I read today online of another one who had this issue, and he was rebuilding indexes.

We also tried working a bit with batch sizes but that didn't really have any noticeable effects.

Evilfish2000 20-Dec-22 5:40am View

Well... "joy". I'm aware that this issue is very much a "only I can fix it". My hope was that I could be pointed into a direction that made sense. I am still working on the issue from my end. I did do some more reading on inserting large amounts of rows. I might need to experiment with my batch sizes it would seem.

Evilfish2000 20-Dec-22 4:58am View

@Graeme_Grant I have looked into this documentation, but extending the timeout just do that. Extends the timeout. We still see the same issue that the server is blocked for hours. Yes we do get a timeout exception, but we want to know why the server stops at the exact same time during our SqlBulkInsert and then just give back to us that timeout error.

Comments by Evilfish2000 (Top 10 by date)