pig-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Erik Krogen (JIRA)" <j...@apache.org>
Subject [jira] [Updated] (PIG-5290) User Cache upload contention can cause job failures
Date Thu, 07 Sep 2017 20:31:00 GMT

     [ https://issues.apache.org/jira/browse/PIG-5290?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]

Erik Krogen updated PIG-5290:
-----------------------------
    Attachment: PIG-5290-1.patch

Thank you for the review [~rohini]!

I see you are right about (3); I missed the line you linked. I can keep it at the old behavior.

For (2), I agree that there is potential for collision during writing temp file, but figured
that since it will fall back to default behavior if it had an error while uploading, this
is not too much of an issue (just an extra JAR upload rather than using cache). I don't understand
the "renaming to final file name" issue. Renames on HDFS overwrite, so if two users upload
the same simultaneously, one of them will simply win (whoever comes second). Let me know if
I am misunderstanding.

Sounds good on (4), thanks for the pointer.

Attaching v1 patch addressing your comments in (3), (4), comment in (1) is no longer relevant.
Will wait on your response about (2).

> User Cache upload contention can cause job failures
> ---------------------------------------------------
>
>                 Key: PIG-5290
>                 URL: https://issues.apache.org/jira/browse/PIG-5290
>             Project: Pig
>          Issue Type: Bug
>    Affects Versions: 0.13.0
>            Reporter: Erik Krogen
>            Assignee: Erik Krogen
>         Attachments: PIG-5290-1.patch, PIG-5290.patch
>
>
> We recently enabled the User Cache (PIG-2672) feature and found that occasionally jobs
would fail because of contention when uploading JARs into the cache. Although the cache is
designed to be fail-safe, i.e. to fall back to normal behavior if anything goes wrong by catching
all {{IOException}}, the portion of code which closes the output stream _is not_ wrapped within
a {{try}} statement and thus an exception during the closing of that stream causes the entire
job to fail. If multiple jobs are attempting to upload the same JAR failure simultaneously,
the contention can cause this close statement to fail.
> The current strategy also has two other flaws. First, consider the scenario where job
A begins uploading jar X. Job B also needs jar X, sees that the file exists, and launches
its tasks. Yet, job A has not yet finished uploading jar X (perhaps it is large). So, the
tasks are localizing a half-completed version of jar X. Second, the original design allowed
for the same JAR (identical contents) to be shared between jobs even if a different name was
used. In PIG-3815, however, this ability was removed, and now JARs are only shared if they
have the same name.
> I propose we solve all of these issues simultaneously by returning to the listStatus
based behavior (used prior to PIG-3815), but filter out entries ending in {{.tmp}}. When uploading,
upload to {{randomNumber.tmp}}, then once the file is completed, do a rename to the original
name of the JAR file. This ensures that incomplete files are never in a location that would
be accessed by other jobs, and the only write operation accessing a shared path is a single
rename operation.
> An alternative design is to use a single canonicalized name for all JAR files (they will
still be unique since they are inside of directories based on their SHA1). Upload to a tmp
file as previously described, then rename to the canonical name. This removes the need to
do a listStatus call; however it will result in classpaths that are human unreadable since
the name of the JAR file has been lost. I think it's worth it from a debugging standpoint
to go with the first design.



--
This message was sent by Atlassian JIRA
(v6.4.14#64029)

Mime
View raw message