RTC: error uploading files during Checkin/Deliver

I encountered the same issue.

In my case, I was trying to update the MS PowerPoint file, which was already in the server.

Hi,

It looks like you are using the RTC Eclipse Client. When I have seen similar error messages, I just reload my workspace and that usually clears this situation up. I believe that you may have encountered a period of "high network traffic" and there was some sort of "network time out" that was picked up by RTC. I am just speaking based on what I have seen. Your experience may be slightly (or greatly) different. Also, someone from the IBM RTC development team may want to weigh in on this with their interpretation of this situation.

Here is my basis for what I am saying:

When I see errors like this, I look in the log on the RTC server. For us, it is located under <tomcat_install_dir>\tomcat\logs (NOTE: this is not the file named "jazz.log"). Only IT Network Services here has access to that log. However this file, named "<tomcat_install_dir>\tomcat\logs\stdout_<date>.log" does have most of the same info that the jazz.log file does.

So, at times around seeing the error in the original post these messages are seen in the log:

=== Begin log info ===

14:59:49,736 ERROR ository.service.internal.RdbRepositoryDataMediator - CRJAZ1038I RuntimeException processing query

BuildRequestQueryModel r = BuildRequestQueryModel.ROOT;
AstQuery q = (AstQuery) IItemQuery.FACTORY.newInstance(r);
q.filter(r.buildDefinitionInstance().buildDefinitionId()._in(new IString[] {q.newStringArg(), q.newStringArg(), q.newStringArg(), q.newStringArg()})._and(r.processed()._eq(0)));
q.orderByAsc(r.created());
com.ibm.team.repository.service.internal.UncheckedRepositoryException: com.ibm.team.repository.common.RetryableDatabaseException: CRJAZ1318I A retryable database error occurred. Database deadlocks and timeouts can occur during periods of high server traffic. You should retry your operation, and contact your server administrator if the problem persists.
Serialization failure
SQL: select t1.ITEM_ID, t1.STATE_ID from BUILD.BUILD_REQUEST t1 where (t1.BUILD_DEFNTN_NSTNC_BLD_DFNTN_D in(?, ?, ?, ?) and (t1.PROCESSED = 0)) order by t1.CREATED asc
SQL Exception #1
SQL Message: Transaction (Process ID 57) was deadlocked on lock resources with another process and has been chosen as the deadlock victim. Rerun the transaction.

=== End of log info ====

Since we have been chosen as the deadlock victim, we just rerun the transaction. In this case, it was a "load" action that was part of a build request and resubmitting the request for a build cleared the issue. This type of situation has also been seen when loading workspaces and running the "load" action again clears the issue up.

As I said earlier, you situation may or may not be exactly what I have experienced, but if it is you now know a way to work through it.

Hope this helps,

- Walter

During an attempt to checkin and deliver a set of files, RTC fails with the following message:

Error during upload

Failed to upload File /ACMEventHandler/test/com/ibm/casemgmt/intgimpl/test/MetadataClassDef.java

Error uploading new contents of /ACMEventHandler/test/com/ibm/casemgmt/intgimpl/test/MetadataClassDef.java

Using Version 2.0.0.2.

Configuration info:

*** Features:
com.ibm.team.concert.feature (2.0.2.I200911251653-2-17J311_23221562) "Rational Team Concert"
org.eclipse.cvs (1.1.2.R342_v20090122-7C79E9x9sLM1t6M9YD7_A7) "Eclipse CVS Client"
org.eclipse.draw2d (3.4.2.v20090218-1145-3317w311_122502441) "Graphical Editing Framework Draw2d"
org.eclipse.emf.codegen (2.4.0.v200902171115) "EMF Code Generation"
org.eclipse.emf.codegen.ecore (2.4.2.v200902171115) "EMF Ecore Code Generator"
org.eclipse.emf.common (2.4.0.v200902171115) "EMF Common"
org.eclipse.emf.ecore (2.4.2.v200902171115) "EMF Ecore"
org.eclipse.emf.ecore.edit (2.4.1.v200902171115) "EMF Ecore Edit"
org.eclipse.emf.edit (2.4.2.v200902171115) "EMF Edit"
org.eclipse.emf.mapping (2.4.0.v200902171115) "EMF Mapping"
org.eclipse.emf.mapping.ecore (2.4.0.v200902171115) "EMF Ecore Mapping"
org.eclipse.gef (3.4.2.v20090218-1145-67728084A56B4I233613552) "Graphical Editing Framework GEF"
org.eclipse.help (1.0.2.R342_v20090122-7r7xEKaEJBZu5oGhP3u6nOmbM9VK) "Help System Base"
org.eclipse.jdt (3.4.2.r342_v20081217-7o7tEAoEEDWEm5HTrKn-svO4BbDI) "Eclipse Java Development Tools"
org.eclipse.pde (3.4.2.R342_v20090122-7T7U1E9imVKz-A8Vz-p_jRS) "PDE"
org.eclipse.platform (3.4.2.R342_v20090122-9I96EiWElHi8lheoJKJIvhM3JfVsYbRrgVIWL) "Eclipse Platform"
org.eclipse.rcp (3.4.200.R342_v20090122-989JESTEbig-SVaL8UJHcYBr4A63) "Eclipse RCP"
org.eclipse.xsd (2.4.2.v200902171115) "XSD - XML Schema Definition Model"


Any thoughts on what may be the cause?

Thanks

I could successfully check-in through the Eclipse Web interface in the same condition.

Hi All,

Could someone resolve this issue without reloading the workspace? My workspace has around 500 MB and reloading that is pretty slow :(

Thanks,

I've encountered the same issue with files modified outside of eclipse, such as Microsoft Word documents. Our users set the editor of "*.docx" filetypes in their preferences to the external Word program. Upon saving the file and attempting to check it in, they get an "error during upload". By highlighting the filename in package explorer or project explorer, and pressing F5 (refresh), this allows the check in to work. Reloading the workspace was not the answer for us as doing such caused users to lose all their local changes, and therefore nothing left to check in.

Hopefully someone on the forum can give you a pointer, but to get help
with an urgent problem, you need to also contact Rational customer
support. They will want to know which client you are using to perform
the checkin (Eclipse, Visual Studio, command line, Web).

And at a minimum, I'd suggest bumping your client up to the current iFix
(iFix 6).

Cheers,
Geoff

On 3/18/2011 10:53 AM, vrcampbell wrote:

This is a ongoing, recurring problem that I have with checking in
changes to a Microsoft Word document. It says: "Check in has
encountered a problem. Checking-in changes to a new change
set". Similar error trying to use an existing change set. I
reload my workspace, try again same error. After 5 tries, it finally
got something to work, only to try it again with yet another change to
the document and have the same problem. By reloading a workspace,
users have lost all of their local changes without warning. At least
an unload first will say that your pending changes will be lost. I
have some angry users expecting me to find a magic solution to this.
So does anyone have a suggestion for me? What is going on here!!!! I
can not emphasize enough the urgency in resolving this. A thought, we
are version 2.0.0.2 iFix2. Did this get any better with version 3?

I had this same problem with jpg image files. I was able to work around it with the following:

Restart client
(Try to rename the file. this caused 'file out of sync' messages)
Refresh the directory/project containing the file.
Checkin/Deliver

Hope this works for you.