We are using Elasticsearch for a few months now, but lately our server's memory went nuts. Indeed, this last one was constantly above 90% and after reading Heap: Sizing and Swapping, I simply limited the initial / maximum size of total heap space to less than fifty percent of the available memory.
It's working great and the memory is now around 70% to 80% without damaging the time of response (or at least, it's not notifiable with my own eyes). I know that my issue may be resolve, but it still bothers me that this issue came up so suddenly. So my question is quite simply: could guys give some advice to improve the memory usage or maybe tell me if something else could be at the origin of this recent change of behavior.
I'm not the person who installed and configured Elasticsearch and so before reviewing the entire documentation, I would prefer to simply ask you. I read a lot about similar issues and some people were talking about that the number of documents may have a similar effect.
Edit: I forgot to talk a bit about our server's configuration which is:
- OS: Windows Server 2012 R2 Datacenter
- Processor: Intel(R) Xeon(R) CPU E5-2660 0 @ 2.20GHz
- Installed memory (RAM): 3.50 GB
- System type: 64-bit Operating System, x64-based processor
I thank you in advance for your response and wish you a great weekend.