I have developed a ETL process to sync a database from an MSSQL instance to Snowflake. Each step in the process is an individual workflow with the process embedded in an orchestration workflow using CReW runners to execute the process in order. most of the processes are run using InDB tools but they are at times needing to stream across DBs.
So my problem is I'm seeing significant performance degradation when comparing a workflow running locally in designer (~1:44 hrs) vs when running on server (~22:54hrs) and am trying to identify possible reasons for the performance difference.
Any ideas on what to check would be appreciated.
It could also be the physical location of the server which gives slow transfer speeds. Just a thought.
Yeah, location of the server would be something also to check.
You'd also want to look at:
I'll add to the list if I can think of anymore things to check too.
@JoeS wrote:Yeah, location of the server would be something also to check.
You'd also want to look at:
- CPU's of the two machines, are they similar?
- RAM similar?
- What's the sort/join RAM setting within Alteryx on each box?
- Are there any firewall rules on the server sending the connections round the houses? (so to speak)
- Same database drivers being used on desktop and server?
I'll add to the list if I can think of anymore things to check too.
I think the issue is the server instance (a dev instance) isn't the size of server that was expected (significantly smaller and smaller than desktop machines). Have a new scaled server being rolled out this week, will confirm if that solves the issue.
User | Count |
---|---|
3 | |
1 | |
1 | |
1 | |
1 |