You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
ℹ️ failed to getLatestVersion: failed to fetch latest subo release: context deadline exceeded
This is due to us synchronously checking for updated subo releases and the server not responding within the deadline. We should make this check even more asynchronous, and remove this notice from production builds (it is confusing, and the user cannot really do anything about it anyway).
@jagger27 proposed to make these checks fully asynchronous and not wait for their results (but show a notice if the last check found a newer version on the next run). In these cases we can also use the failure state (now unbound from the short completion deadline) as an actual metric we collect and send back via telemetry to report potential breakage.
The text was updated successfully, but these errors were encountered:
ℹ️ failed to getLatestVersion: failed to fetch latest subo release: GET https://api.github.com/repos/suborbital/subo/releases/latest: 403 API rate limit exceeded for 174.115.111.169. (But here's the good news: Authenticated requests get a higher rate limit. Check out the documentation for more details.) [rate reset in 7m20s]
Yeah we should probably at least have a file that has a timestamp of last check and we don't do it more than every couple hours (that would go well with @jagger27's proposal of doing this async as you can also just serialize into the file the latest found version, which could be compared with the current version on next run)
Subo intermittently throws this notice:
This is due to us synchronously checking for updated
subo
releases and the server not responding within the deadline. We should make this check even more asynchronous, and remove this notice from production builds (it is confusing, and the user cannot really do anything about it anyway).@jagger27 proposed to make these checks fully asynchronous and not wait for their results (but show a notice if the last check found a newer version on the next run). In these cases we can also use the failure state (now unbound from the short completion deadline) as an actual metric we collect and send back via telemetry to report potential breakage.
The text was updated successfully, but these errors were encountered: