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
In #1578, there was a network stall that could only be worked around with a timeout. It would be better if we could detect the stall itself, rather than relying on an overall timeout.
In today's system, we can only detect that a request overall takes too long, and abort it. Ideally, we could have a separate timeout for stalls. For example, if timeout is 60s, and stallTimeout is 5s, it would mean that the request overall could take up to 60s, but that if we ever went 5s with no progress at all, the request would be aborted anyway.
The text was updated successfully, but these errors were encountered:
In #1578, there was a network stall that could only be worked around with a timeout. It would be better if we could detect the stall itself, rather than relying on an overall timeout.
In today's system, we can only detect that a request overall takes too long, and abort it. Ideally, we could have a separate timeout for stalls. For example, if
timeout
is 60s, andstallTimeout
is 5s, it would mean that the request overall could take up to 60s, but that if we ever went 5s with no progress at all, the request would be aborted anyway.The text was updated successfully, but these errors were encountered: