Strange timeout issues with Alfresco 4

Options
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
05-24-2012 07:22 AM
I am successfully retrieving documents (amongst other operations) from Alfresco using the DotCMIS library in C#. I have no issues when connecting to our older Alfresco 3.3 server.
When using Alfresco 4, however, the DotCMIS calls (fairly reliably) stop responding after I've fetched *any* two (sometimes three) documents, even the same one each time (I have created the same documents on both servers, to rule out anything peculiar with the documents themselves.) In this state, any subsequent call eventually raises a Timeout exception, although sometimes they will just magically start working again, most of the time the only resolution is to restart my application. It's just peculiar because this problem does not occur when I am using the older Alfresco server, I can fetch as many documents as I like with no issues.
I don't think the stack trace is very useful:
The call I am making is
I have checked the logs on the Alfresco server and there is nothing at all related to the failed calls.
Has anybody else seen this? I can't rule out that my code is at fault, but there must be some key difference between the versions of Alfresco that cause this behaviour if this isn't a problem with Alfresco's CMIS implementation…
When using Alfresco 4, however, the DotCMIS calls (fairly reliably) stop responding after I've fetched *any* two (sometimes three) documents, even the same one each time (I have created the same documents on both servers, to rule out anything peculiar with the documents themselves.) In this state, any subsequent call eventually raises a Timeout exception, although sometimes they will just magically start working again, most of the time the only resolution is to restart my application. It's just peculiar because this problem does not occur when I am using the older Alfresco server, I can fetch as many documents as I like with no issues.
I don't think the stack trace is very useful:
at DotCMIS.Binding.AtomPub.AbstractAtomPubService.Read(UrlBuilder url) in chemistry-dotcmis-0.4-src\src\binding\atompub\atompub.cs:line 436 at DotCMIS.Binding.AtomPub.AbstractAtomPubService.GetObjectInternal(String repositoryId, IdentifierType idOrPath, String objectIdOrPath, Nullable`1 returnVersion, String filter, Nullable`1 includeAllowableActions, Nullable`1 includeRelationships, String renditionFilter, Nullable`1 includePolicyIds, Nullable`1 includeAcl, IExtensionsData extension) in chemistry-dotcmis-0.4-src\src\binding\atompub\atompub.cs:line 742 at DotCMIS.Binding.AtomPub.ObjectService.GetObject(String repositoryId, String objectId, String filter, Nullable`1 includeAllowableActions, Nullable`1 includeRelationships, String renditionFilter, Nullable`1 includePolicyIds, Nullable`1 includeAcl, IExtensionsData extension) in chemistry-dotcmis-0.4-src\src\binding\atompub\atompub.cs:line 1712 at DotCMIS.Client.Impl.Session.GetObject(String objectId, IOperationContext context) in chemistry-dotcmis-0.4-src\src\client\client-impl.cs:line 532 at DotCMIS.Client.Impl.Session.GetObject(IObjectId objectId, IOperationContext context) in chemistry-dotcmis-0.4-src\src\client\client-impl.cs:line 500 at DotCMIS.Client.Impl.Session.GetObject(IObjectId objectId) in chemistry-dotcmis-0.4-src\src\client\client-impl.cs:line 490 at GetDocumentFromID(String document_ID) in CMISDocumentManager.cs:line 533
The call I am making is
ob = session.GetObject(id) as IDocument;
I have checked the logs on the Alfresco server and there is nothing at all related to the failed calls.
Has anybody else seen this? I can't rule out that my code is at fault, but there must be some key difference between the versions of Alfresco that cause this behaviour if this isn't a problem with Alfresco's CMIS implementation…
Labels:
- Labels:
-
Archive
5 REPLIES 5
Options
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
05-30-2012 06:03 PM
Are you using the deprecated web script based CMIS implementation (has a service URL of /alfresco/s/api/cmis or similar) or the new OpenCMIS based CMIS implementation (has a service URL of /alfresco/cmisatom or /alfresco/cmisws)?
The new OpenCMIS implementation did not exist until Alfresco 4.
There are definitely implementation differences between the two.
Jeff
The new OpenCMIS implementation did not exist until Alfresco 4.
There are definitely implementation differences between the two.
Jeff

Options
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
05-31-2012 12:21 PM
Hi Jeff,
I'm using the OpenCMIS implementation at /alfresco/cmisatom (although we might temporarily change this due to another issue, so I will investigate if this problem persists once this change has been made.)
Thanks,
Katie.
I'm using the OpenCMIS implementation at /alfresco/cmisatom (although we might temporarily change this due to another issue, so I will investigate if this problem persists once this change has been made.)
Thanks,
Katie.
Options
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
08-08-2012 06:24 PM
Hi,
we are having same issues as Katie!
For better understanding: we are doing migration from FileNet P8 to Alfresco through custom app, based on CMIS. We can do maybe several dozen files before timeout.
We are using Alfresco EE 4.0.2 with latest DotCMIS (0.4) from Apache Chemistry.
Binding is AtomPub. There is nothing in server logs.
After first timeout all subsequent operations also timeout.
Any new insights?
Best Regards,
Robert
we are having same issues as Katie!
For better understanding: we are doing migration from FileNet P8 to Alfresco through custom app, based on CMIS. We can do maybe several dozen files before timeout.
We are using Alfresco EE 4.0.2 with latest DotCMIS (0.4) from Apache Chemistry.
Binding is AtomPub. There is nothing in server logs.
After first timeout all subsequent operations also timeout.
Any new insights?
Best Regards,
Robert

Options
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
08-16-2012 06:47 AM
Further to this, I have exactly the same issue with another DMS, so this does not appear to be a problem with Alfresco specifically. I have logged an issue with the Chemistry project to see if they can make any sense of this: https://issues.apache.org/jira/browse/CMIS-559

Options
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
12-18-2015 01:11 PM
Katie, Did you ever resolve this?
