> must set to avoid all shards go to one node? Did you mean primary shards?
The difference between primary shards and replicas is very small (almost a flag). > node1 has more LOAD then the node0 What do you by load? If you are talking about OS load, you are probably doing the bulk request to node1 and he is sending information to node0. So yes, node1 has more load than node0. If you make the bulk request to node0, you will see (I trully hope it) that you get the opposite LOAD. Notice that it does not depend on the primary or replica thing. Regards, Pablo On Tuesday, July 22, 2014 10:25:59 PM UTC-3, 天行 wrote: > > When i restart the nodes, i find all primary shards go to the node1, and > all replica shards go to the node0. > So, when i bulk write to the es, i find the node1 has more LOAD then the > node0. > I want to ask whether some thing i must set to avoid all shards go to one > node? > > The follow is the monitor for the two node when i bulk write to es: > > <https://lh3.googleusercontent.com/-sXUtxqGpQng/U88OPdOy6FI/AAAAAAAAAmU/HjvZahKTdoE/s1600/es1.jpg> > > <https://lh6.googleusercontent.com/-i43TeIgmDDk/U88Op7qr-qI/AAAAAAAAAmc/fZQx-G_n-Jk/s1600/es2.jpg> > > <https://lh3.googleusercontent.com/-HxyBmw2CEHU/U88O7234MxI/AAAAAAAAAmk/waP5Q59QF44/s1600/es3.jpg> > > > -- You received this message because you are subscribed to the Google Groups "elasticsearch" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To view this discussion on the web visit https://groups.google.com/d/msgid/elasticsearch/a0ba135f-8a75-48d3-9497-8417ea668936%40googlegroups.com. For more options, visit https://groups.google.com/d/optout.
