[
https://issues.apache.org/jira/browse/HDDS-17481?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=18132314#comment-18132314
]
Anurag Parvatikar commented on HDDS-17481:
------------------------------------------
One scenario to be documented :-
AllocateBlock: If the bucket policy changes before a subsequent
{{allocateBlock}} made from an older client during the write, then some blocks
of the key will not respect the {{StoragePolicy}} of the key.
Example:
# A Bucket with storagePolicy=HOT exists.
# An older client not support storagePolicy writes a key, key1:
## Since key1 has no policy passed by the client, key1 assumes the bucket
policy = HOT.
## Blocks are pre-allocated with storagePolicy is HOT.
# While data is being written to the pre-allocated blocks of key1, bucket's
storagePolicy is changed to COLD.
# Now the write path of key1 makes another allocateBlock call - Since there is
no storagePolicy in the keyArgs, bucket policy, COLD, is used when allocating
the new blocks.
# Resulting key, key1 will have some blocks in HOT - SSD tier and some blocks
in COLD - ARCHIVE tier.
> [DOC] Document behaviour for edge cases.
> ----------------------------------------
>
> Key: HDDS-17481
> URL: https://issues.apache.org/jira/browse/HDDS-17481
> Project: Apache Ozone
> Issue Type: Sub-task
> Reporter: Anurag Parvatikar
> Assignee: Anurag Parvatikar
> Priority: Major
>
> This Jira is to track edge cases and their expected behaviour found during
> implementation of the Storage Policy feature.
--
This message was sent by Atlassian Jira
(v10.3.26#10030026)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
