Elasticsearch getting very slow

[shikhar_singh (shikhar)]

what i observed:

1. The kibana dashboards were not getting refreshed.
2. Navigating the kibana UI was very slow.

what i observed on the server side .

what i did as a work around

after clearing the cache my ES/KIbana came back to normal

Can you please help me know the RCA for this , why ES was unable to use the buffered/cache memory ?

[jbur]

Linux uses buff/cache to accelerate disk operations, and normally it will not take memory away from processes that need memory. Have any non-standard configuration parameters been implemented that would change the way Linux manages memory?

Also, are you having storage performance issues? If you run top, what do you see for %wa while the issue is occurring?

While you're searching for the root cause, you could try to tune the cache to avoid having to manually flush or drop the cache:
https://access.redhat.com/documentation/en-us/red_hat_enterprise_linux/5/html/tuning_and_optimizing_red_hat_enterprise_linux_for_oracle_9i_and_10g_databases/sect-oracle_9i_and_10g_tuning_guide-memory_usage_and_page_cache-tuning_the_page_cache

I hope this helps. Please let us know what you find!