Swapping can come handy if a cluster node reaches its maximum memory
capacity and, instead of failing with an out-of-memory exception, you can
keep it alive by swapping out data before the cluster is scaled out. Once
you add a new cluster node, the data will be rebalanced and the first node
will fr
Hello!
I recommend using native persistence in this case.
Regards,
--
Ilya Kasnacheev
пт, 31 янв. 2020 г. в 09:07, ashishb888 :
> Thank you Evgenii!
>
> I can understand that enabling swap may decrease the performance.
> Keeping in that mind I wanted to see what actually happens after enablin
Thank you Evgenii!
I can understand that enabling swap may decrease the performance.
Keeping in that mind I wanted to see what actually happens after enabling
swap.
For that I need to know how to enable swap on a specific node without
affecting other nodes in the cluster.
BR,
Ashish
--
Sent fr
Hi,
First of all, I don't think running 2 server nodes on the same physical
machine is a good idea. They will compete for resources and they will need
to support 2 internal processes(like discovery and communication), which
will have an overhead. It's better to have only one here, or have 2 smalle
I have,
-4 nodes 2 clients & 2 severs. A & B are the clients and X & Y are the
servers.
-32 GB RAM
-2 data regions default & xyz
-A & B takes 2 GB RAM each. X takes 10 GB & Y takes 15 GB but may grow
to 25 GB
-X uses default region and Y uses xyz region
All the above nodes run
