HDDS-5483. Validate block file length during put block

[sadanand48]

What changes were proposed in this pull request?

Validate block file length during put block .

What is the link to the Apache JIRA

https://issues.apache.org/jira/browse/HDDS-5483

How was this patch tested?

Only a validation check.

[smengcl]

Under what conditions could this be violated?

[sadanand48]

@smengcl This is just a safety check. The ozone streaming write path will not use the current ratisAsyncApi which ensures sequential transactions i.e putBlock would always occur after writeChunk. To be on a safer side, this check is added. We may remove this after the complete streaming implementation if this check is redundant
cc : @mukul1987

[smengcl]

lgtm

[ferhui]

Maybe this should rebase on HDDS-4454

[mukul1987]

+1, LGTM

[szetszwo]

@sadanand48 , @mukul1987 , Should this be merged to the master branch instead the HDDS-4454 branch? This dose not look like specific to HDDS-4454.

[szetszwo]

@sadanand48 , @mukul1987 if there is no objection, I will merge this to the master branch. Thanks.

[mukul1987]

@szetszwo, sure lets do that

[szetszwo]

Thanks for the response. Just have merged this to the master branch.

[adoroszlai]

I think layout-specific code would be better added to each ChunkManager implementation instead of such == and isAssignableFrom checks. Take a look at finishWriteChunks() and shutdown().

[adoroszlai]

Should this be merged to the master branch instead the HDDS-4454 branch?

Although I don't expect any problems for this specific change, I would prefer going via PR for master next time, to make sure all checks pass.

[szetszwo]

@adoroszlai , it is a good idea to have a PR. I will revert the commit from the master.

[szetszwo]

commit a595d06450d6788237bd459a1ac4f36fe5582992 (HEAD -> master, origin/master, origin/HEAD)
Author: Tsz-Wo Nicholas Sze <szetszwo@apache.org>
Date:   Thu Aug 19 10:53:26 2021 +0800

    Revert "HDDS-5483. Validate block file length during put block (#2463)"
    
    This reverts commit 375f3c4f02a46e12fcfc1acc127ecdc5badd359d.

[szetszwo]

@sadanand48 , as suggested by @adoroszlai , please submit a PR for the master branch. Similar to other JIRAs, we will rebase this to the HDDS-4454 branch after this has been committed to the master branch. Thanks.