I have a 5-node cluster. The Kudu master server is installed in name node 1, and three tablet servers are installed in data node 1, 2, and 3. After running for two days, the tablet server on dn 2 dies. Other tablet servers are still working fine. We want to investigate the failing tablet server. Please advise where to find the kudu log file. Thanks!
I found the log in dn2: /var/log/kudu
F0521 07:13:04.131083 65713 hybrid_clock.cc:339] Check failed: _s.ok() unable to get current time with error bound: Service unavailable: clock error estimate (20016888us) too high (clock considered unsynchronized by the kernel)
Please advise how to fix this problem.
I think I bit late to respond .
But Basically kudu will not start or fail if the server’s clock is not synchronized using NTP.
One way to fix is to run on the tablet server host - sudo service ntp restart or stop/ start .
And you can fire ntp status .