This message was deleted.
# helpdesk
s
This message was deleted.
r
Likewise, I’d like to better what the
EgressStatus
enum actually corresponds to. What status is returned if the underlying egress instance fails? What status is returned if streaming to the RTMP endpoint fails (looks like Egress Failed)?
Also, when doing API error handling, at least from the js sdk, the error type is
AxiosError
. Is that guaranteed to be stable?
d
that's great feedback. the enums values can be found here, but I agree it's pretty confusing. you should be able to reference the type from the JS SDK tho. Regarding API error handling, it's not guaranteed that it'll be
AxiosError
. We are actually moving from axios to fetch for better compatibility with other JS runtimes
r
Yeah so I did look through all the proto files but it’s kinda unclear what event would cause the corresponding error status
For the api error handling, it’d be useful if as part of that change, the sdk wrapped the underlying axios or fetch exception into a livekit specific one
👍 1
And kinda on that same note, it’d be nice to have in the documentation what errors calls would return in the course of normal operation, so that we know in which cases the calls should be retired. For example, when I was testing yesterday, I found that calling stop on an already stopped egress raised a twirp error of 503, which usually is a transient and retriable error code, but is not in this case, which is why some additional documentation would be useful.
Although this is all obviously just kinda regarding error handling. In the course of normal operation the apis and documentation are quite straightforward and easy to use 🙂