i guess that he wants to terminate the execution if it does not finish online. In mysql they have query hint like: SELECT /*+ MAX_EXECUTION_TIME(1000) */ --in milliseconds * FROM tableSELECT /*+ MAX_EXECUTION_TIME(1000) */ --in milliseconds * FROM table
since workers only lives for 30 seconds per fetch invocation, i guess he want it failed rather than worker eviction to raise a reasonable error to the user, or to do something else
It would be nice if the d1 dashboard showed the performance stats for each query! It’s probably already coming through the API right? My small db is already using a lot more rows read that I would like. Would like some easier ways to investigate queries.
Yeah I used the insights. It’s very useful, thank you for adding it. I learned one of my queries read 1 million rows in the past 24h. I just added some new indexes. Just thinking now I want to experiment with one off versions of that query to see if there is any improvement but it’s still hard to measure. Wow, I’ll be looking forward to the dash improvements!
We will probably rebuild a product with 5 million of users on it. It’s a chat base application which use an external provider for the chat. So the db only handle some informations about users and it’s mostly use for a community feature with a list of all users and a search on it. Do you think D1 could be a good fit for that us case? (I am a bit afraid of reaching the size limit)
For the search part probably (I was also thinking using a worker ai for the search part with vectorize) but I cannot put everything in it no? I would still need something for the community part and some other features