Hi, Eric It is defined in doc/connection-backoff.md > <https://github.com/grpc/grpc/blob/master/doc/connection-backoff.md>.
Actually, I'm curious how the specification is integrated into the CQ logic. For the auto reconnection, my concern is if I want to implement my own retry logic, since the reconnection is actually controlled in the background thread which poll the cq. The user retry logic sometimes are not working, since grpc cannot make sure every time user retry a rpc will auto reconnect the channel (e.g. at the retry time, the channel state is also TRANS_FAILURE). Thanks, Yihao 在 2017年9月21日星期四 UTC-7上午11:31:03,Eric Anderson写道: > > 在 2017年9月21日星期四 UTC-7上午11:09:49,yihao yang写道: >> >> I guess the default wait_for_ready option is just to wait the exponential >> backoff reconnect until it timeouts and will not issue the channel >> reconnect immediately. Am I right? > > > Yes, the wait_for_ready says you'd rather wait than get a failed RPC. So > it does not fail the RPC if the connection is known-bad (which is when > we're in exponential backoff). It waits until either 1) a connection is > established or 2) the RPC's deadline is exceeded. > > On Thu, Sep 21, 2017 at 11:11 AM, yihao yang <yangyi...@gmail.com > <javascript:>> wrote: > >> BTW, I just curious about the implementation of the channel reconnect. >> And I cannot find any docs about it. >> > > It is defined in doc/connection-backoff.md > <https://github.com/grpc/grpc/blob/master/doc/connection-backoff.md>. > -- You received this message because you are subscribed to the Google Groups "grpc.io" group. To unsubscribe from this group and stop receiving emails from it, send an email to grpc-io+unsubscr...@googlegroups.com. To post to this group, send email to grpc-io@googlegroups.com. Visit this group at https://groups.google.com/group/grpc-io. To view this discussion on the web visit https://groups.google.com/d/msgid/grpc-io/2542cd84-1904-42d3-8966-05e234461c68%40googlegroups.com. For more options, visit https://groups.google.com/d/optout.