It is taking in excess of 10 seconds sometimes a lot more) to load the app logs for a brand new app with only a handful (less than 10 originally) entries. It just sits there with the blue spinner.
This is the first time I’ve noticed this for a relatively empty log file on a new app.
I’ve seen logs access being painfully slow for a couple of weeks now.
I tend to only use them when creating APIs, and this is the first one I’ve done in months.
It’s agony. 3 minutes is the record so far to load 27 lines.
I created my own logger app, so I can log from both Anvil apps and any other kind of tools and see the aggregate log in one spot.
When I need to access the log very often during development and the official log is too slow, I either use my logger app, or (if it’s a temporary thing) I add a log table to the app being developed and use that as a log.
App logs have been (practically) unusable for some time.
It used to be fast to see the latest entries, but searching was often failing.
Now searching works, but it’s so slow that it’s unusable, and it’s slow even looking at the latest entries.
I was going to make the API handler do nothing except accept the call, then push it all out to an uplink based handler which can log locally.
I seem to remember someone making the standard python logger work with anvil, but I might be wrong.
There are more elegant ways to do it using the standard logger, but every time I try I hit conflicts with libraries that don’t play nice with the standard logger, and I end up with missing or duplicated log entries. So I made my own.
Using the uplink could be slow. I would (well I did) create a Logger app with an http endpoint so it can be used by Excel macros and other tools. The app has one form with a repeating panel and some search options to show the content of the Log table.
Any app can share the Log table and add rows to it.
My Logger is both faster and easier to search than the Anvil app logger.
The main reason why I made my Logger was to log from other tools, but now I’m using also to replace the broken Anvil logger.
+1 on this…
Edit: 3 hours later and BOOM. Lightning fast! Thanks!
Update: You might like to check out the App Logs again now 
The logs are back!!!
Wow, Curiosity: what was the issue because this is lightening fast?
Hope I can apply the same logic to my applications.
Now that’s a LOT better!
Cheers!
Roughly, “tickle Postgres’s tummy juuust right so that it uses the nice helpful index we have, rather than trawling every un-coalesced event in the log database of which there are Quite A Few”.