Shared storage is definitely not the way to go, that would defeat the strength of ES. And I think you will also run into issues if they access the same folder.
How much storage you need per node depends on your ES config (and of course your input volume, which you already know). You can shard your indices and you can also replicate them. Sharding does not have an impact on storage needs, but replicas do. If you start with 2 nodes and have 1 replica and 2 shards all of your data will be mirrored across the two nodes. 2 shards with 2 nodes means ES will automatically distribute the shards on both nodes - one each. This is only for performance. If you have zero replicas then your 5TB will be divided by the number of nodes. On Tuesday, November 25, 2014 9:11:58 PM UTC+1, Robert Breault wrote: > > I am new to Graylog2 and have a questions regarding the setup of multiple > ES servers. > > We will be running 2 ES nodes to start and it is an understand that each > node needs its own storage or can I use a shared storage? > > We will be collecting around 100 GB a day as we are trying to scale out > the disk do I need lets say I plan on having 5 TB of data do I need 5 TB > for each node or can I have just 5 TB and have that shared between both > nodes? > > > Thank you > -- You received this message because you are subscribed to the Google Groups "graylog2" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. For more options, visit https://groups.google.com/d/optout.
