Hi @meredydd thanks for clarifying this!
We still seem to be getting a weird issue this morning where we’re getting the error in relation to our uplinked code.
anvil.server.UplinkDisconnected: Uplink disconnected
This only happens for 1 function (out of about 20) even though it is marked as ‘registered’ on the Anvil logs. The behavior isn’t persistent too, it seems to fail every 1/2 tries.
We’ve also had a couple of complete outages this morning.
yes I am getting this a lot too
[silently cheering on the Anvil team]
(not seeing any issues myself, to be clear)
I really appreciate the hard work, and I honestly don’t think leaving anvil is an option for me. The productivity upswing is too much (unless gpt obsoletes everything and I just tell it what it do).
Having said that, I think it would be extremely beneficial for the community to have a system put in place for large scale issues. This has been mentioned before, and we can all set up our own personal alert’s, but it doesn’t give us any talking points to give our respective client (whether that is our management or customers). It also provides some relief getting any feedback on the issue being acknowledged by the anvil team.
A simple
“We are investigating the issue”
“We have found the issue and are working towards a solution”
“We are solving the issue”
“Issue has been solved”
We would have to define what classifies a large scale issue, but I think that could be defined, better than I, by the community.
I agree. openai has a good SMS alert system.
MAINLY I would appreciate the notice so i know the difference between something I’ve messed up and something that I know anvil are working on.
I’ve been getting this the last day or so, using my IDE-specific uplink codes. (Production’s Uplink codes and programs seem unaffected.) This is Uplink code that has been working for months, so it’s unlikely to be something on our end.
In case this helps with diagnostics, I’m including some details here.
Just switched from anvil-uplink v0.4.1 to v0.4.2, still getting the error. The funny thing is, I can see my Uplink programs connected and running in my 2nd monitor. They “think” they’re still connected!
After killing and restarting the affected Uplink program, it says it successfully connects, but an Uplink call to it still raises anvil.server.UplinkDisconnected.
Unplugging the network cable from the computer, waiting a minute, and plugging it back in, results in some recovery messages from the Uplink program, but it does reconnect. For the next minute or so, calls to that Uplink program succeed. After that, they fail.
Rebooting the entire computer, and restarting the Uplink programs, seems to make the next Uplink call succeed. But that’s probably just a longer disconnect.
I’m using Python v3.7.5 under Windows 10 Professional. (v3.7 is intentional, to have least surprise when testing app’s server- and browser-side code locally.) I could go up to v3.7.9. Later editions of v3.7 require that I compile Python myself. Not an option.
Anvil-uplink is listed as compatible up to v3.9. v3.9.13 can be installed without having to compile Python.
I’m currently having uplink disconnected errors on an app that hasn’t been changed in quite a while. Rebooted the server the uplink is running on, still have issues. I’m not seeing any errors or disconnects on the server running the uplink, just in the browser and app logs.
Likely related:
Thank you for the detailed and non-sugarcoated response. All I can say is that Anvil is literally the bee’s knees.
I’m still getting the same issue unfortunately and I worry about our customers still facing issues in the morning after the long bank holiday weekend here in the UK.
I am also getting Uplink disconnected errors on an old app that has not changed in a very long time.
I am still stuck with both errors 
:
Update: Over the weekend we experienced some lingering issues due to incorrect system state from last Friday (distributed systems are hard!), which should have been resolved this morning with a hard reset.
Can those experiencing issues subsequent to my last post please confirm whether these are still going on now (Tuesday afternoon, UK time)? If so, these are new phenomena unrelated to this thread (to our knoweldge), and we’ll want to get on those - so please include app IDs and any hex codes you see.
Both errors are still blocking my app from functioning.
Not sure what do you mean by:
Hi @clinical,
What I meant there was a request to identify the app that’s having the problems (the easiest way to do this is the ID that appears in the URl when you’re editing an app - eg https://anvil.works/build/apps/CQNXEFWWK376VXTP), dates/times of examples when the error occurred, and ideally a link+instructions for how to reproduce the error, so that we can look at our logs and investigate the issue, and if necessary reproduce it ourselves for testing.
All of my apps are working normally again 
Errors from this afternoon:
app ID: APL36MR5C2TIJHO5
RuntimeUnavailableError: Remote server not available (downlink_call). Please contact support@anvil.works.: sessions: 5BPJDGWYCJBWCHSBIOZDQ2ZKCSWPL2WB, CIYDDSSBPH762BV6L6FMGK4FD7RLEUGX, T6FTCEJOUC6A5NEMA3NYDW24EHEGGUB4 TimeoutError: Server code took too long: sessions HGPF4A4SVP3Y5QHUYLCUELBPEGXULSWW, R7VDI5UDPH3EUL3UJNUMQEYGTB727SJERuntimeUnavailableError: Downlink disconnected: 1c2ecb8dde3a: session HJQ2KHKZPZ3GJ2KL2KDG3TPJXJVJWNKCLooking at the logs, I can’t see any pattern with these, so was not able to reproduce them.
Sure, I sent it through the email.
by the way, my code is very simple:
@anvil.server.callable
def get_data():
return [i['project_name'] for i in app_tables.projects.search(selected=True)]
It is a table issue.
here is another error:
anvil.server.RuntimeUnavailableError: Remote server not available (:exec_call). Please contact support@anvil.works.Update (just now): it seems that
anvil.server.RuntimeUnavailableErrors
stoped, however the load is very slow, and I am still getting
anvil.server.TimeoutError: Server code took too long
?