[
https://issues.apache.org/jira/browse/CRAIL-109?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17055825#comment-17055825
]
Jonas Pfefferle commented on CRAIL-109:
---------------------------------------
Crail does not aim to store data persistent. Please take a look at
[http://crail.incubator.apache.org/overview/]
to understand the purpose of Crail. In short: Crail is designed for temporary
data, which does not require to be made persistent as ones the job is finished
the data typically is deleted.
That said there is a way to make data in Crail persistent by using the NVMf
datatier. Also you need to enable logging capabilities on the namenode to make
metadata information persistent.
> RDMAstorage tier seem to do nothing about data persistency
> ----------------------------------------------------------
>
> Key: CRAIL-109
> URL: https://issues.apache.org/jira/browse/CRAIL-109
> Project: Apache Crail
> Issue Type: Bug
> Affects Versions: 1.2
> Environment: Ubuntu18.04
> Reporter: zhubohong
> Priority: Major
> Labels: persistence, rdma
> Original Estimate: 1h
> Remaining Estimate: 1h
>
> It seems that rdma storage tier is based on mapped file , and did nothing to
> persist data.
> or there lack of clear and simple docs to configure persist feature
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
