|
I have a team I'm working with that is running 10 SQL Anywhere databases. They are running this on Ubuntu 18.04 SQL Anywhere is version 17.0.11.7312 They are wanting to introduce replication so they can use it to move data to their Azure environment. Long story short, one of the databases is 300GB is size. Currently they have been extracting for almost 24 hours and it looks like they are about 80% done. Their machine has 8 cores, and for the most part it has just been using 1 core, with the occasional blip up to 2 cores. It never goes beyond that. Is there any way to speed this process up? They are more or less throwing their hands up thinking that extractions taking this long would cause issues since they are running isolation level 3. This is not a disk space issue from what I can tell. They have about 100TB available on the database partition. Any thoughts or suggestions on what might speed this process up would be greatly appreciated. Jeff Gibson Nashville, TN Additional info - We are watching the core usage and SQL Anywhere just jumped up to 8 cores, but it's only using about 12% of each core. But it literally just started using this. |
|
I would not suggest you run dbxtract on an in use production consolidated database, but rather that you run the extract on a copy of the consolidated database.
Reg |
When you say "they are wanting to introduce replication" am I correct in assuming that you mean SQL Remote?
When you say "extracting" do you mean running dbxtract against the database?
I'm going to assume yes and answer the question, and will delete the answer if my assumptions are wrong.
Are there complicated article definitions contained, say with several joins based on subscribe by values? That may slow down extraction heavily.
My apologies Reg. Yes. They are using SQL Remote. And this is only one way replications. So they are doing sends only from production and receive only at the two other environments.
It is not. It's basically a select all saying send everything that happens.