[
https://issues.apache.org/jira/browse/HADOOP-13665?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15959762#comment-15959762
]
Wei-Chiu Chuang commented on HADOOP-13665:
------------------------------------------
[~lewuathe] thanks again for your quick turn-around!
The v10 patch looks very close.
A few nits
* IO_ERASURECODE_CODEC_XOR_RAWCODERS_DEFAULT should also support native
fallback, and its value should be
org.apache.hadoop.io.erasurecode.rawcoder.NativeXORRawErasureCoderFactory,org.apache.hadoop.io.erasurecode.rawcoder.XORRawErasureCoderFactory.
io.erasurecode.codec.xor.rawcoders in core-default should be updated as well.
* In getRawCoders(), io.erasurecode.codec. can be replaced by
IO_ERASURECODE_CODEC.
* The private method getRawCoderFactNameFromCodec() is no longer used, and its
use is replaced by createRawEncoderWithFallback() and
createRawDecoderWithFallback. It should be removed if I understand it
correctly. The associated variables such as
IO_ERASURECODE_CODEC_RS_LEGACY_RAWCODER_KEY should also be removed as well
because they are not used.
* In createRawEncoderWithFallback() and createRawDecoderWithFallback(), I think
it's better to print the exception so that we know what goes wrong. So instead
of
{code}
LOG.warn("Failed to create raw erasure encoder " + factName +
", fallback to next codec if possible");
{code}, do
{code}
LOG.warn("Failed to create raw erasure encoder " + factName +
", fallback to next codec if possible", e);
{code}
* We should also update the documentation HDFSErasureCoding.md, but I prefer to
do that in a separate jira.
> Erasure Coding codec should support fallback coder
> --------------------------------------------------
>
> Key: HADOOP-13665
> URL: https://issues.apache.org/jira/browse/HADOOP-13665
> Project: Hadoop Common
> Issue Type: Sub-task
> Components: io
> Reporter: Wei-Chiu Chuang
> Assignee: Kai Sasaki
> Priority: Blocker
> Labels: hdfs-ec-3.0-must-do
> Attachments: HADOOP-13665.01.patch, HADOOP-13665.02.patch,
> HADOOP-13665.03.patch, HADOOP-13665.04.patch, HADOOP-13665.05.patch,
> HADOOP-13665.06.patch, HADOOP-13665.07.patch, HADOOP-13665.08.patch,
> HADOOP-13665.09.patch, HADOOP-13665.10.patch
>
>
> The current EC codec supports a single coder only (by default pure Java
> implementation). If the native coder is specified but is unavailable, it
> should fallback to pure Java implementation.
> One possible solution is to follow the convention of existing Hadoop native
> codec, such as transport encryption (see {{CryptoCodec.java}}). It supports
> fallback by specifying two or multiple coders as the value of property, and
> loads coders in order.
--
This message was sent by Atlassian JIRA
(v6.3.15#6346)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
