|
|
Niels Charlier created an issue |
|
Issue Type: |
|
|---|---|
|
Assignee: |
|
|
Components: |
JDBCConfig, JDBCStore |
|
Created: |
10/Jul/18 3:05 PM |
|
Priority: |
|
|
Reporter: |
|
The hazelcast cache is a huge drain on resources (very slow to do catalog queries, and MASSIVE data transfers between nodes). It’s probably even faster to just query the database than the cache. See discussion: I would turn it off and let have each node have its own cache, jdbcconfig already invalidates the cache based on events. We do need to do something similar for jdbcstore which also uses the cache. |
|
|
|
Get Jira notifications on your phone! Download the Jira Cloud app for Android or iOS |
|
|
This message was sent by Atlassian Jira (v1001.0.0-SNAPSHOT#100088-sha1:bf99e1e) |
|
|
|