One problem that our customers appear to be running into is checking in projects. We do have a bug that appears to leave projects checked out when you wouldn't expect them to be. But I suspect that in many cases it may be that the architectural changes to 2007 have caused some of this confusion - and the default view of the queue may well hide the true cause of the problem.
So - check out these different scenarios to see if this covers your issues:-
UPDATE: Thanks to Ed Hanna and Brian Kennemer for bringing it to my attention - but it appears that closing Project using the X in the top right can lead to check-in issues. So best practice would be first to save and ensure that any communication with the server has completed (look in the status area after saving) and then using File, Close.
Technorati Tags: Project Server 2007
Having all the same problems listed above. Have also tried all of the suggestions above. Same problems. The extra little fun issue I have is that Proj Pro errors when I try to access the Local Cache Cleanup option. So I can't try cleaning the local cache. Any suggestions?
You may need to take the heavy approach and deleting the cache from the OS rather than through the UI. It may be the local cache causing the errors. It can be found at:-
Vista - C:\Users\brismith\AppData\Roaming\Microsoft\MS Project
XP etc. C:\Documents and Settings\brismith\Application Data\Microsoft\MS Project
Unless you have pointed it somewhere else.
I can confirm that the we are experiencing the exact same check-in problems as mentioned above. We have all adopted the 'do not touch the x' to close projects approach, but we still get the problem.
The only way to resolve this is to remove the local cache, via UI or OS. We therefore lose work, but at least we can work on the plan.
Brian, do you think this could somehow be related to fractions in the work / remaining work columns? We have found that the problems above happen in certain instances only, which made us look into the data that is being passed in and out of the plans (via timesheets)
The only shot in the dark that we have is that somehow a fraction is being logged and this somehow affects the local cache. Sounds weird, but there is just no other explanation for this strange behaviour.
Furthermore, I was wondering where we could find an official response from MS regarding the problems mentioned above?
I am also facing the same problem. Tried all the workaround mentioned in the earlier comment. Finally what I have observed is that the Queue is not getting processed at all. What could be the reason for queue getting stuck ?
So does the workaround of opening read-only and then closing again not work for you? I am not aware of a relationship to the update of fractional work - but if you get a good workaround I would love to hear it. As for an official response from Microsoft - we are aware of this issue and hope to have a fix for this in SP1 when it is shipped.
One reason may be the queue is not actually started. You should see two queue processes (assuming you have just one shared services provider) and if you only see one in Task manager then I would suggest restarting the service. It may not have started if SQL Server could not be contacted when the machine started up. I have seen this in virtual environments. For other queue troubleshooting tips look elsewhere on this blog. It is rare that the queue is actually stuck - but individual jobs can block other activity - just a case of finding what those jobs are.
Having the same issue and this thread help us get though. Just to confirm with you, if we see that the Active Cache Status is 'Succeeded' and we then clear the cache, do we still stand a chance of losing data?
A force check-in of the project prior to this also showed a success status in the queue service.
As long as all queue jobs related to that project and client are successful then you can clear the cache. The cache is used for loading the projects too - assuming it has not been updated on the server - so it can be useful to keep the projects there, as it can make loading faster if you are working on a slow network.
I've tried the above method, but I still can't fix my problem, which I encountered frequently these days. And my issue is that, I can't force checkin a project(supposed project A here). I have checked the queue and canceled the relevant job; also, I try read-only open A and close it, but A still cannot be forced checkin. Is there any other idea to force checkin project A? Can I resolve it by clear the check-out ststus in database?
By the way, project server SP1 is now available, I hope many bugs encountered now will be fixed in this shippment.
Though not exact, we have experienced simliar issues. They seemed to be resolved with the following procedure in Pro:
1) Save the project
2) Publish (if that's intended)
3) File>Close, responding "Yes" to checkin.
If the last step is skipped by closing the window with the "X" or File>Exit (Alt>f>X), the project remains Checked-out.
Hope this helps most, if not all.
No problems installing the SPs for Server or WSS 3.0 and the SP1 for Project Pro seemed to correct issues on my laptop but two other PMs are still experiencing the "Check-in Pending" when trying to open a project hours after the project was checked-in and shows checked in on the server side. Any ideas? Their local cache settings have been set at the minimun 50mb but their still having the same issue. They're running the same OS and configuration I am and I'm at a loss at this point.
I have had a problem where a Job has got stuck on 93% complete, I have isolated the date that appears to be at issue, but the problem is the back up of the job occurred later that day. It hangs on the same 93% as the original file which looks like it was shut and publish cancelled at this status.
I can save the file, but not publish it and we have a bunch of time recorded against this project.
What is the best procedure to recover the file without losing the data?, and what are the options. I have perused the blog and tried a number of the steps.
Still having the "checkin pending" issue.
Server side shows the project is checked in. Only the client thinks it is still checkin pending.
Clearing the local cache solves that problem.
Found this issue when I experienced a cross-up between my Admin account checking out a project and being unable to check the project back in. Tried the "force" method, tried to delete cache, tried to create Project Server Account for use within Project (profiles) for the Admin account. Now I have a check-in status of Cancelled and cannot save the project, open/edit the project or any other manipulation. I am hard stuck w/out the ability to even delete the project and re-publish. H E L P ???
Are you running SP1 or later? Can anyone else see/open the project? What error do you get? You probably need a support incident to helpget you back from this bad place.