Gargi-jais11 commented on code in PR #9318:
URL: https://github.com/apache/ozone/pull/9318#discussion_r2537640113
##########
hadoop-hdds/framework/src/main/java/org/apache/hadoop/hdds/utils/HddsServerUtil.java:
##########
@@ -422,7 +422,7 @@ public static Collection<String>
getOzoneDatanodeRatisDirectory(
if (rawLocations.isEmpty()) {
rawLocations = new ArrayList<>(1);
- rawLocations.add(ServerUtils.getDefaultRatisDirectory(conf));
+ rawLocations.add(ServerUtils.getDefaultRatisDirectory(conf, "dn"));
Review Comment:
After further analysis I found that :
In production environments, having multiple DataNodes on the same host is
NOT recommended for exactly the reasons you stated:
**Single point of failure** - If the host fails, all DNs fail
**Resource contention** - CPU, memory, disk I/O, network bandwidth all shared
**Defeats distributed storage purpose** - No fault tolerance benefit
**Poor performance** - Multiple DNs competing for same resources
Production documentation confirms this:
"Bare metal machines are recommended for optimal performance"
Typical production: 1 DataNode per physical host
So I don't think this case will happen.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
