[
https://issues.apache.org/jira/browse/HADOOP-19604?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=18012797#comment-18012797
]
ASF GitHub Bot commented on HADOOP-19604:
-----------------------------------------
anmolanmol1234 commented on code in PR #7853:
URL: https://github.com/apache/hadoop/pull/7853#discussion_r2262094329
##########
hadoop-tools/hadoop-azure/src/main/java/org/apache/hadoop/fs/azurebfs/services/AbfsBlobClient.java:
##########
@@ -1103,8 +1107,21 @@ public AbfsRestOperation flush(byte[] buffer,
AbfsRestOperation op1 = getPathStatus(path, true, tracingContext,
contextEncryptionAdapter);
String metadataMd5 = op1.getResult().getResponseHeader(CONTENT_MD5);
- if (blobMd5 != null && !blobMd5.equals(metadataMd5)) {
- throw ex;
+ /*
+ * Validate the response by comparing the server's MD5 metadata
against either:
+ * 1. The full blob content MD5 (if full blob checksum validation is
enabled), or
+ * 2. The full block ID buffer MD5 (fallback if blob checksum
validation is disabled)
+ */
+ if (getAbfsConfiguration().isFullBlobChecksumValidationEnabled() &&
blobMd5 != null) {
+ // Full blob content MD5 mismatch — integrity check failed
+ if (!blobMd5.equals(metadataMd5)) {
Review Comment:
The way in which the 2 md5 are computed are different, blobMd5 computes the
entire blob data's md5 while md5Hash is used only for the md5 computation of
the list of blockId's and hence we are maintaining 2 separate variables.
> ABFS: Fix WASB ABFS compatibility issues
> ----------------------------------------
>
> Key: HADOOP-19604
> URL: https://issues.apache.org/jira/browse/HADOOP-19604
> Project: Hadoop Common
> Issue Type: Sub-task
> Affects Versions: 3.4.1
> Reporter: Anmol Asrani
> Assignee: Anmol Asrani
> Priority: Major
> Labels: pull-request-available
> Fix For: 3.4.1
>
>
> Fix WASB ABFS compatibility issues. Fix issues such as:-
> # BlockId computation to be consistent across clients for PutBlock and
> PutBlockList
> # Restrict url encoding of certain json metadata during setXAttr calls.
> # Maintain the md5 hash of whole block to validate data integrity during
> flush.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: common-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: common-issues-h...@hadoop.apache.org
