snowcrumble commented on issue #6123: Improve recover time from SPOF URL: https://github.com/apache/pulsar/issues/6123#issuecomment-581780621 > * ZooKeeper session timeout. default 30s. Before the session expires, the ownership of topics still belongs to the failed broker. > * Client keep-alive probes. The client sends periodic probes in the connection to detect stale TCP connections. The default time is 1min. In Java client it's possible to configure the interval: https://pulsar.apache.org/api/client/org/apache/pulsar/client/api/ClientBuilder.html#keepAliveInterval-int-java.util.concurrent.TimeUnit- Hi, I have tried to set broker configuration as below: ``` zooKeeperSessionTimeoutMillis: "10000" zooKeeperOperationTimeoutSeconds: "10" ``` But the situation has not improved, first my situation is on connecting moment, so it's nothing to do with keep-alive, then I found it finally cost about 1 min on cpp code `async_resolve`, it had no logs within `Created connection for ...` and `Failed to establish connection: No route to host`. May it be improved to add a timeout on `async_resolve` op?
---------------------------------------------------------------- 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: [email protected] With regards, Apache Git Services
