hadoop-common-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Steve Loughran (JIRA)" <j...@apache.org>
Subject [jira] [Updated] (HADOOP-9361) Strictly define the expected behavior of filesystem APIs and write tests to verify compliance
Date Fri, 17 Jan 2014 17:54:22 GMT

     [ https://issues.apache.org/jira/browse/HADOOP-9361?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]

Steve Loughran updated HADOOP-9361:
-----------------------------------

    Attachment: HADOOP-9361-006.patch

Patch with spec consistent with almost all HDFS behaviour, s3 and localfs set up to throw
tighter exceptions on some failures, and to throw EOFException on seek(-negative). 

h2. LocalFS behaviours to resolve

1. attempting to mkdir over an existing file returns false instead of raising an exception.

propose: raise an exception. Nobody ever checks the return code from mkdirs after all -so
we should uprate it to be on a par with HDFS.

2. you can seek on an stream valid after a close()

options:
*  fix
* ignore on the basis reads or writes will fail when attempted


3. if you rename a file over an existing file, the operation succeeds. This is what bash does.

propose: document as the less preferred option; relax test to permit with a warn

h2. HDFS contract test behaviours

1. if you open a stream for append, rename the file and then do the append, the old filename
remains.

Propose: specify the outcome as "undefined"

2. if you attempt to rename a file that doesn't exist to a path in the same directory, it
returns false, rather than raising a FileNotFoundException

I'm assuming here that the dest path is being checked before the source. I'd consider this
an error

3. delete("/", true) returns false and doesn't delete anything

I've documented this as valid behaviour, and noted it is what HDFS does.

> Strictly define the expected behavior of filesystem APIs and write tests to verify compliance
> ---------------------------------------------------------------------------------------------
>
>                 Key: HADOOP-9361
>                 URL: https://issues.apache.org/jira/browse/HADOOP-9361
>             Project: Hadoop Common
>          Issue Type: Improvement
>          Components: fs, test
>    Affects Versions: 3.0.0, 2.2.0
>            Reporter: Steve Loughran
>            Assignee: Steve Loughran
>         Attachments: HADOOP-9361-001.patch, HADOOP-9361-002.patch, HADOOP-9361-003.patch,
HADOOP-9361-004.patch, HADOOP-9361-005.patch, HADOOP-9361-006.patch
>
>
> {{FileSystem}} and {{FileContract}} aren't tested rigorously enough -while HDFS gets
tested downstream, other filesystems, such as blobstore bindings, don't.
> The only tests that are common are those of {{FileSystemContractTestBase}}, which HADOOP-9258
shows is incomplete.
> I propose 
> # writing more tests which clarify expected behavior
> # testing operations in the interface being in their own JUnit4 test classes, instead
of one big test suite. 
> # Having each FS declare via a properties file what behaviors they offer, such as atomic-rename,
atomic-delete, umask, immediate-consistency -test methods can downgrade to skipped test cases
if a feature is missing.



--
This message was sent by Atlassian JIRA
(v6.1.5#6160)

Mime
View raw message