[
https://issues.apache.org/jira/browse/HDDS-3707?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17127887#comment-17127887
]
maobaolong edited comment on HDDS-3707 at 6/10/20, 7:40 AM:
------------------------------------------------------------
[~elek] [~arp] Thank you for your reply, i agree with that UUID collision
chance can be very-very low, but if we can supply a way(API) to generate the id
from a center service which can be named UUIDGeneratServer, by default, we can
use the local UUID generator, if someone configure a UUIDGeneratServer, i can
be better if we can get new UUID from UUIDGeneratServer.
was (Author: maobaolong):
[~elek] [~arp] Thank you for your reply, i agree with that UUID collision
change can be very-very low, but if we can supply a way(API) to generate the id
from a center service which can be named UUIDGeneratServer, by default, we can
use the local UUID generator, if someone configure a UUIDGeneratServer, i can
be better if we can get new UUID from UUIDGeneratServer.
> UUID can be non unique for a huge samples
> -----------------------------------------
>
> Key: HDDS-3707
> URL: https://issues.apache.org/jira/browse/HDDS-3707
> Project: Hadoop Distributed Data Store
> Issue Type: Bug
> Components: Ozone Datanode, Ozone Manager, SCM
> Affects Versions: 0.7.0
> Reporter: maobaolong
> Priority: Minor
> Labels: Triaged
>
> Now, we have used UUID as id for many places, for example, DataNodeId,
> pipelineId. I believe that it should be pretty less chance to met collision,
> but, if met the collision, we are in trouble.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
