manifoldcf-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Karl Wright (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (CONNECTORS-1532) Moving a file outside of the job's Paths is not the same as deleting it
Date Wed, 19 Sep 2018 18:26:00 GMT

    [ https://issues.apache.org/jira/browse/CONNECTORS-1532?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16621006#comment-16621006
] 

Karl Wright commented on CONNECTORS-1532:
-----------------------------------------

Whenever you click "Start" on such jobs, seeding is done.  Documents are located based on
their folder affinity and added to the queue.  The documents that were added are then processed.
 At the end of the job, when document processing is complete, documents that were not discovered
at seeding time are removed from the index.

There are numerous integration tests in ManifoldCF that verify this basic sequence.  The Documentum
connector has almost nothing to do with this process unless your operation of "moving the
document out of the folder" is not in fact removing the document from whatever folder structure
metadata Documentum maintains, and the document is still discoverable with the same DQL query
as before.  The final deletion is a function of the framework, not of any of the connectors.

The only difference I know of is that the Documentum Connector uses MODEL_ADD, which is not
very common in connectors, while the tests use connectors that use MODEL_ADD_CHANGE.  This
should not have any effect on the cleanup cycle but I can trivially confirm that that is true.

If I cannot find any difference, the next step would be for me to modify the FileSystem connector
(used in the tests extensively) to use MODEL_ADD instead.  Then I can see if the tests still
pass.  If they do, then I'd be forced to conclude that you aren't changing the Documentum
folder in the way you think you are.


> Moving a file outside of the job's Paths is not the same as deleting it
> -----------------------------------------------------------------------
>
>                 Key: CONNECTORS-1532
>                 URL: https://issues.apache.org/jira/browse/CONNECTORS-1532
>             Project: ManifoldCF
>          Issue Type: Bug
>          Components: Documentum connector
>    Affects Versions: ManifoldCF 2.10
>         Environment: Manifold 2.10 patched for #1512, #1517
>            Reporter: James Thomas
>            Assignee: Karl Wright
>            Priority: Major
>         Attachments: 2018-09-19_1758.png
>
>
> If I have a MF job which is connecting a specific folder, F, in Documentum to a File
System output then:
> 1. deleting files in Documentum shows them as zero size in the file system
> 2. moving files out of F does not remove them or zero them in the file system
> Note that moving a file from another folder (which the job is not looking at) to F has
the same effect as adding it to F by e.g. importing it in DM or POSTing it to DM via the REST
interface.
> Intuitively, I expect that moving a file out of the "view" of the Documentum connector
would have the same effect on the File System as deleting it. (My model here is of MF synchronising
content between the Paths (DM) and the Output Path (File System) that I have specified in
the job.)
> Starting point, I have run the MF job to fetch a bunch of files from a folder - call
it F - in DM (i.e. I have configured Paths in the job to be F). This is what 'ls -l' on the
file system looks like:
> {code:java}
> -rw-r--r--. 1 root i2e  12541 Sep 19 07:21 drl?versionLabel=CURRENT&objectId=090000018000f7c0
> -rw-r--r--. 1 root i2e     26 Sep 19 07:21 drl?versionLabel=CURRENT&objectId=090000018000f7be
> -rw-r--r--. 1 root i2e  85772 Sep 19 07:21 drl?versionLabel=CURRENT&objectId=090000018000f7c7
> -rw-r--r--. 1 root i2e   8790 Sep 19 07:21 drl?versionLabel=CURRENT&objectId=090000018000f7c2
> -rw-r--r--. 1 root i2e 101888 Sep 19 07:21 drl?versionLabel=CURRENT&objectId=090000018000f7c3
> -rw-r--r--. 1 root i2e  32783 Sep 19 07:21 drl?versionLabel=CURRENT&objectId=090000018000f7c4
> -rw-r--r--. 1 root i2e  23040 Sep 19 07:22 drl?versionLabel=CURRENT&objectId=090000018000f7c1
> -rw-r--r--. 1 root i2e  26112 Sep 19 07:22 drl?versionLabel=CURRENT&objectId=090000018000f7bf{code}
> In DM, I delete one of the files in F and it shows as zero size, and the modification
date has changed:
> {code:java}
> -rw-r--r--. 1 root i2e  12541 Sep 19 07:21 drl?versionLabel=CURRENT&objectId=090000018000f7c0
> -rw-r--r--. 1 root i2e     26 Sep 19 07:21 drl?versionLabel=CURRENT&objectId=090000018000f7be
> -rw-r--r--. 1 root i2e   8790 Sep 19 07:21 drl?versionLabel=CURRENT&objectId=090000018000f7c2
> -rw-r--r--. 1 root i2e 101888 Sep 19 07:21 drl?versionLabel=CURRENT&objectId=090000018000f7c3
> -rw-r--r--. 1 root i2e  32783 Sep 19 07:21 drl?versionLabel=CURRENT&objectId=090000018000f7c4
> -rw-r--r--. 1 root i2e  23040 Sep 19 07:22 drl?versionLabel=CURRENT&objectId=090000018000f7c1
> -rw-r--r--. 1 root i2e  26112 Sep 19 07:22 drl?versionLabel=CURRENT&objectId=090000018000f7bf
> -rw-r--r--. 1 root i2e      0 Sep 19 07:23 drl?versionLabel=CURRENT&objectId=090000018000f7c7{code}
> In DM, I move a file from F to another folder. (Right click, add to clipboard, go to
new folder, Edit> Move here). 
> The file shows as modified (07:25), but is still apparently in F (i.e. in the Path my
MF job is looking at):
> {code:java}
> -rw-r--r--. 1 root i2e  12541 Sep 19 07:21 drl?versionLabel=CURRENT&objectId=090000018000f7c0
> -rw-r--r--. 1 root i2e     26 Sep 19 07:21 drl?versionLabel=CURRENT&objectId=090000018000f7be
> -rw-r--r--. 1 root i2e   8790 Sep 19 07:21 drl?versionLabel=CURRENT&objectId=090000018000f7c2
> -rw-r--r--. 1 root i2e 101888 Sep 19 07:21 drl?versionLabel=CURRENT&objectId=090000018000f7c3
> -rw-r--r--. 1 root i2e  23040 Sep 19 07:22 drl?versionLabel=CURRENT&objectId=090000018000f7c1
> -rw-r--r--. 1 root i2e  26112 Sep 19 07:22 drl?versionLabel=CURRENT&objectId=090000018000f7bf
> -rw-r--r--. 1 root i2e      0 Sep 19 07:23 drl?versionLabel=CURRENT&objectId=090000018000f7c7
> -rw-r--r--. 1 root i2e  32783 Sep 19 07:25 drl?versionLabel=CURRENT&objectId=090000018000f7c4{code}
> In DM, I move a file from another folder to F and it shows up with the timestamp of the
move (07:28):
> {code:java}
> -rw-r--r--. 1 root i2e  12541 Sep 19 07:21 drl?versionLabel=CURRENT&objectId=090000018000f7c0
> -rw-r--r--. 1 root i2e     26 Sep 19 07:21 drl?versionLabel=CURRENT&objectId=090000018000f7be
> -rw-r--r--. 1 root i2e   8790 Sep 19 07:21 drl?versionLabel=CURRENT&objectId=090000018000f7c2
> -rw-r--r--. 1 root i2e 101888 Sep 19 07:21 drl?versionLabel=CURRENT&objectId=090000018000f7c3
> -rw-r--r--. 1 root i2e  23040 Sep 19 07:22 drl?versionLabel=CURRENT&objectId=090000018000f7c1
> -rw-r--r--. 1 root i2e  26112 Sep 19 07:22 drl?versionLabel=CURRENT&objectId=090000018000f7bf
> -rw-r--r--. 1 root i2e      0 Sep 19 07:23 drl?versionLabel=CURRENT&objectId=090000018000f7c7
> -rw-r--r--. 1 root i2e  32783 Sep 19 07:25 drl?versionLabel=CURRENT&objectId=090000018000f7c4
> -rw-r--r--. 1 root i2e 191513 Sep 19 07:28 drl?versionLabel=CURRENT&objectId=09000001800045b9{code}
> But if I immediately move it out in DM then, again, the timestamp (07:30) alters but
the file apparently remains:
> {code:java}
> -rw-r--r--. 1 root i2e  12541 Sep 19 07:21 drl?versionLabel=CURRENT&objectId=090000018000f7c0
> -rw-r--r--. 1 root i2e     26 Sep 19 07:21 drl?versionLabel=CURRENT&objectId=090000018000f7be
> -rw-r--r--. 1 root i2e   8790 Sep 19 07:21 drl?versionLabel=CURRENT&objectId=090000018000f7c2
> -rw-r--r--. 1 root i2e 101888 Sep 19 07:21 drl?versionLabel=CURRENT&objectId=090000018000f7c3
> -rw-r--r--. 1 root i2e  23040 Sep 19 07:22 drl?versionLabel=CURRENT&objectId=090000018000f7c1
> -rw-r--r--. 1 root i2e  26112 Sep 19 07:22 drl?versionLabel=CURRENT&objectId=090000018000f7bf
> -rw-r--r--. 1 root i2e      0 Sep 19 07:23 drl?versionLabel=CURRENT&objectId=090000018000f7c7
> -rw-r--r--. 1 root i2e  32783 Sep 19 07:25 drl?versionLabel=CURRENT&objectId=090000018000f7c4
> -rw-r--r--. 1 root i2e 191513 Sep 19 07:30 drl?versionLabel=CURRENT&objectId=09000001800045b9{code}
> In DM, I now delete all visible content in F. The files that were moved out of F, and
are not visible in F in DM, remain on the file system:
> {code:java}
> -rw-r--r--. 1 root i2e      0 Sep 19 07:23 drl?versionLabel=CURRENT&objectId=090000018000f7c7
> -rw-r--r--. 1 root i2e  32783 Sep 19 07:25 drl?versionLabel=CURRENT&objectId=090000018000f7c4
> -rw-r--r--. 1 root i2e 191513 Sep 19 07:30 drl?versionLabel=CURRENT&objectId=09000001800045b9
> -rw-r--r--. 1 root i2e      0 Sep 19 07:31 drl?versionLabel=CURRENT&objectId=090000018000f7c2
> -rw-r--r--. 1 root i2e      0 Sep 19 07:31 drl?versionLabel=CURRENT&objectId=090000018000f7be
> -rw-r--r--. 1 root i2e      0 Sep 19 07:31 drl?versionLabel=CURRENT&objectId=090000018000f7c0
> -rw-r--r--. 1 root i2e      0 Sep 19 07:31 drl?versionLabel=CURRENT&objectId=090000018000f7c1
> -rw-r--r--. 1 root i2e      0 Sep 19 07:31 drl?versionLabel=CURRENT&objectId=090000018000f7bf
> -rw-r--r--. 1 root i2e      0 Sep 19 07:31 drl?versionLabel=CURRENT&objectId=090000018000f7c3{code}



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)

Mime
View raw message