caiconghui opened a new issue #4023:
URL: https://github.com/apache/incubator-doris/issues/4023
Although we have max_running_txn_num_per_db config to prevent excessive
system load in fe, but frequent retry may also cause excessive system load in
fe, in our production environment, we see many logs like
`retrying call frontend service,
address=TNetworkAddress(hostname=10.118.32.48, port=9020), reason=THRIFT_EAGAIN
(timed out)` within a short time, and find high system load in fe, some times
cause fe core, the main reason is that there are already many backlogs in fe,
which cause be thrift client rpc timeout, but be thrift client retry
immediately without any breaks, which cause things more worse, finally make
avalanche disaster in fe.
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscr...@doris.apache.org
For additional commands, e-mail: commits-h...@doris.apache.org
