Closed sandra708 closed 9 years ago
Hi Sandra,
Some good news, some bad news.
The good news: the query actually is working! The query was not actually dead yet.
The bad news: the query was just running for a really long time. It was making great progress. In the most recent step, the code was running for 45 minutes (!!!) and then finished on 70/72 workers, and on the remaining two it is still going.
My guess is that there are memory issues and that it's related to the problem we discussed last week with the max/min and the fact that this ends up meaning we have 8 copies of the reachability table all joined together at the same time.
The soon-to-be-good news is that I have implemented the back-end changes necessary to make this query faster, the bad news is that they're not available from the web yet..
For now, I have killed and restarted the cluster, but probably we need to not resubmit this particular query for now.
@dhalperi can we close this?
Ohyeah
— Sent from my phone
On Wed, Mar 25, 2015 at 2:48 PM, Brandon Myers notifications@github.com wrote:
@dhalperi can we close this?
Reply to this email directly or view it on GitHub: https://github.com/uwescience/myria-web/issues/188#issuecomment-86228592
Can you stop the currently running query (25739)? I think it's not likely to return data if it hasn't yet, and then I can resubmit.