转发：consult some questions about riak,thank you
270917674 at qq.com
Fri Jan 13 11:21:28 EST 2017
Dear basho team:
I encountered a strange question, we use riak/riak-cs build a cluster, the cluster has 7 nodes, DB of riak KV is multi(level & bitcask), earch node has 16G RAM.total_leveldb_mem_percent=50.
today we found one node crash down but without crash log, and erlang log has something like this:
terminate called after throwing an instance of 'std::bad_akkic' wath(): std::bad_alloc
（riak kv version：2.0.5）
[os_mon] memory supervisor port(memsup):Erlang has closed;
[os_mon] cpu supervisor port(cpu_sup):Erlang has close.
And we restart this node and it can work properly again, use the (free -m) command to view the memory, +-buffered/cache: used=2000Mb available<14000Mb. and memory status of other nodes are: used 6000Mb~7000Mb.
can you help me analysis of why the node crash down?
and i found other problem on other cluster. under normal conditions，the memory of used was 4000Mb~5000Mb,but
every 20 minutes will rise once, about doubled, and then returned to normal, why?
thank you very much!
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the riak-users