Brian Smith's Microsoft Project Support Blog

Microsoft Project Conference 2012: What do you want to hear?

Brian Smith - MSFT — Mon, 30 Jan 2012 17:38:39 GMT

Myself and Adrian Jenkins are attending and presenting at the Project Conference 2012, March 19th to the 22nd in Phoenix Arizona and will be delivering a couple of the sessions. The sessions are planned to be one about the February 2012 Cumulative Update for Project and Project Server, and the other one will be a guide to troubleshooting Project Server – looking at some of the tools we use in support to help identify issues. But if there are some specific scenarios or tools that you are especially interested in then we would be delighted to consider these for inclusion in the sessions. The conference is your chance to hear great things from our partners, customers and of course Microsoft – and here is your chance to let me and Adrian know exactly what you’d like to hear from us.

Looking forward to meeting friends old and new in Phoenix!

December 2011 CU Webcast: Questions and Answers

Brian Smith - MSFT — Mon, 30 Jan 2012 16:52:00 GMT

Sorry for the delay in getting this posting out. Vacation and an unusual amount of snow are a couple of the excuses, and if you missed the webcast then you can view here - https://www.livemeeting.com/cc/mseventsbmo/view?id=1032493964&role=attend&pw=5A2EBE71

Also if you are thinking of watching this or future TechNet webcasts then I’ll remind you that you will need the Live Meeting client install which can be obtained from here http://office.microsoft.com/en-us/downloads/CD010254990.aspx. – but on to the questions – thanks for all of you who attended and for the great questions!:

Q: What is the risk of applying the December Cumulative Update on the server but not on the client? Is it highly recommended to always keep them in sync, or is it acceptable to have some clients that are not patched to the same CU as the server?

A: There is no dependency between the client and server, and although to be fully patched you should update both there is no risk (beyond not having some of the fixes) if you update the client or server independently. Once you have patched the client it is also worth ensuring no un-patched clients can connect by use of the Server Setting to set the minimum client patch level that can connect (2010 only)

Q: Ref 30900 – the fix relating to custom calendars - is that custom enterprise calendar, or custom calendar in template, or both?

A: In this case, it is whether or not the template has the custom calendar associated with it. For example, suppose you have a custom calendar that is the base calendar for some tasks within your project. It is in this case that the PSI method would fail as it attempted to reconcile the calendar.

Q: The Save Error 9000 is a very generic message. Can you repeat the exact issue for which this fix was designed?

A: The scenario is this. As you save your project from Project Professional, you see the 9000 error and your project does not get saved. We found about four different conditions where duplicate IDs or objects such as views could appear and it's because of these duplicates that the save process failed. We've added fixes on both the client and server to address these various conditions. As Brian states, there could possibly be other conditions that could lead to this issue and so time will tell. As mentioned – you don’t need to install both server and client fixes – they are independent – but to get full protection you will.

Q: In regards to Timesheets in Single Entry Mode, if a timesheet is submitted and accepted with actuals against 3 projects, and later one of those projects is deleted, and then someone deletes that timesheet and attempts to create a replacement timesheet, currently that results in an error. (The timesheet line items from the deleted projects still remain and seem to stop the new timesheet from being created.) Does this CU address this? The issue is that there is an "orphan" timesheet line record.

A: With the December CU the practice of deleting timesheets should be reviewed within your organization as the behavior changes – but it should not give an error. When the timesheet is re-created it will re-load the current tasks with any current actual work that has been accepted. The deleted project will no longer appear. The specific error mentioned isn’t familiar – so there may still be a risk of it occurring – possibly related to pending actions?

Q: You spoke about graphical indicators disappearing occasionally after changes have been made to their enterprise field definitions. Are there other graphical indicator problems.... such as enterprise fields showing incorrect indicators after changes made to their field definitions? I experienced some problems in one of my deployments but I believe the issue was a wrong indicator not a missing indicator. The problem seems to have gone away when I created new fields in the production site. Forum postings gave me the impression the December CU would address this. Any comments on the status? Are all of these graphical indicator problems resolved now?

A: We have certainly addressed a number of issues with graphical indicators – it could be possible to see the wrong indicator if a condition of a formula calculation gives an unexpected result and no indicator (or an existing indicator) meets the resultant condition. We’d be keen to hear of any reproducible steps that are still giving bad results.

Q: I know that for Project Server 2007 there was a tool (P12ACTool) that could be run against the environment that would check for standard issues behind the scenes and provide steps for resolving those issues. This was used in my past as a preventative step to keeping the environment up to par. Is there a similar tool for Project Server 2010?

A: There is a tool that we use for 2010 (P14ACTool). It is used on support incidents under the direction of a support engineer and not publically available. I does require some interpretation of the results and can give false positives which is why we like to be involved with the customer using the tool. We do appreciate that some partners make extensive use of the tool with their customers and are reviewing a version of the tool that we could make more available – watch this space. We are also investing a lot in diagnostics at Microsoft so expect some new automated packages sent to you from CSS to aid in collection of data.

Q: What is the status of the fix for the 2010 client issue with "Save and Send" where the data is not actually saved to the Server?

A: Although I remember hearing about a condition like this I tried some tests after the webcast and I was not able to reproduce or find a bug in our database. I’d welcome any repro steps so we can address this if it is still an issue.

Q: We have a couple of different 2010 client machines running the Oct and Dec CU, and the links between tasks are vanishing and then reappearing, resource usage view is jumping around between resources when modifying planned time phased actual work, and indicators in the status line (at the bottom) are vanishing - is this a known issue? (not seen it prior to August client CU for 2010)

A: We have a hotfix request in for the resource usage view jumping – likely to be fixed in April CU – not familiar with the other issues – repro steps would aid our investigation here – and confirmed this was more a display problem than the links really going away. Possibly a UI handle problem?

Q: Should you be able to update Actual Start in the Project Center views in edit mode

A: This has not been exposed for editing in the web – possibly because of the knock on effect on the task and assignment start dates.

Q: Are either of you going to be at the Project Conference?

A: We can now confirm that yes, Adrian and I will be at the Project Conference in Phoenix, along with others from the support team: our Manager Larry Block and escalation colleague Aik Chen. Also look out for my next blog post asking what you would like us to talk about at the conference.

Q: This is a general Project Server config question: Is there a way to report time a the high level and not at task level because we just record total time of a project? Meaning entering the time

A: Reporting in the timesheet is possible at the top level

Q: Not really project related but I recently downloaded a few of the past webcasts in wma format. But I can't sync any of these to my Zune. Any suggestions?

A: I tried this by downloading and syncing to my Windows Phone 7 and also this didn’t work. I’ll see if I can find a way of doing this – I’m sure using Expression Encoder would allow re-encoding in a suitable format. ***Update - http://support.microsoft.com/kb/932614 gives good instructions for doing this with Windows Movie Maker, Expression Encoder 4 or Windows Media Encoder - just trying it now... ***

Q: What is the correct way to submit a Design Change Request? Is there a website we can go to directly?

A: No web site – usually giving feedback through your account team, or to a support engineer will get through to the right people.

Project Server 2010: Operation still processing–don’t shoot the messenger

Brian Smith - MSFT — Thu, 29 Dec 2011 21:57:00 GMT

Decided to make this post as we have been getting quite a few questions about this dialog – Operation still processing.

It can occur in various places within Project Web App and in my case I have made it appear in Project Center by creating a new project based on an Enterprise Project Type (called “Large”) that is using a project template which has around 2500 tasks. In this case the full text says

“The Create operation has been queued, but is taking longer than expected to process. To check the status of the creation queue job, visit the My Queue Jobs page. Once the job has completed, the new Large will be listed in the Project Center.
Click OK to close the project and return to Project Center.”

This isn’t an error – I wouldn’t even go so far as to say it is a warning – merely an informational dialog that we programmed to appear in cases where whatever you are doing is taking a little while to happen – usually around a minute or so. It could be that something has gone wrong – more likely it is just taking a while – so have a look in the queue. Possibly by the time you have read the message the job has completed. It could be that you are creating a quite large plan from a template – or perhaps there are other jobs in the queue that mean yours is just waiting. So this shouldn’t be treated as an error – unless when looking in the queue you find a bigger problem – like I did when writing this posting…. Possibly more on that in another posting if a reboot of my abused server doesn’t make my issues go away.

*** Update - Turns out in my attempted repro of this condition and using a large template I exposed a bug. If you have a template of around 1600 tasks or bigger (the exact failure point will depend on other factors - could be as high as 4500) then you may experience a stack overflow when this hits the queue in the Project Create stage. It will sit there processing for ever - and if you look in the ULS logs you will see evidence that the queue has restarted -

12/30/2011 07:46:48.20 Microsoft.Office.Project.Server (0x1EF4) 0x2798 Project Server General 8zdb High [SERVICE] ProjectQueueService14: ProcessWatcher signaled
12/30/2011 07:46:48.20 Microsoft.Office.Project.Server (0x1EF4) 0x2798 Project Server General 8zdk High [SERVICE] ProjectQueueService14: ProcessWatcher restarting service for f0bf6473-d0bd-4442-8d28-65466f8f6dc0 PID: 8160

In the application event logs you will see an event 1000 Application Error followed by one or more Windows Error Reporting event id 1001:

Faulting application name: Microsoft.Office.Project.Server.Queuing.exe, version: 14.0.6015.1000, time stamp: 0x4d0ffe25

Faulting module name: KERNELBASE.dll, version: 6.1.7601.17651, time stamp: 0x4e21213c

Exception code: 0xe053534f

Fault offset: 0x000000000000cacd

Faulting process id: 0x%9

Faulting application start time: 0x%10

Faulting application path: %11

Faulting module path: %12

Report Id: %13

When the queue restarts it will just attempt the project create again, and after around 4 minutes you will see the same errors repeated. Best open a support incident to let us help you recover from this if you run into it. So far I am not aware of any customers running in to this. ***

Project 2010: Events now raised by changes in Project Information dialog

Brian Smith - MSFT — Tue, 20 Dec 2011 19:15:00 GMT

This will be the first of a couple of postings about some changes in behavior that you might see once you install the December Cumulative updates. This one is specifically about the client, and the Project Information dialog. In the past if you changed a field value within the Project Information dialog then this would not raise the ProjectBeforeTaskChange or ProjectDeforeTaskChange2 events. The values in Project Information relate to the project summary task (Task 0) and in Project 2007 these events would fire regardless of where the change was made. Now, with the December CU these events are available to you. However, there are a couple of things you need to be aware of:

1. Firstly the event will fire once for the complete dialog, so unlike if you were changing values in the Project Summary task you may need to allow for and have your users prepared for multiple pop-ups (depending what you code does) once they click OK on the Project Information dialog.

2. If you cancel then you will lose all your changes – so make sure users are aware they may need to go back and check the values if for some reason your code rejected one of the changes. For example if you popped up a dialog for each change and accepted a couple then rejected one the values that had been accepted get cancelled too.

3. There appears to be a bug which we are investigating whereby the Project Department field will always raise the event even if it has not been changed.

I know that we didn’t have the issue with the Department field in 2007 – I will double check the behavior of the second bullet and see if that has also changed since 2007 – and if that should be considered a bug too.

*** Update - looks like this wasn't the behavior in 2007 - you could cancel any of the individual events and not roll back other changes - so I'd consider this a bug in 2010. No promises on any fix dates though.

One workaround to avoid cancelling if you thought the Department field had been changed would be to check if the value has actually changed – something like:

If ((Field = FieldNameToFieldConstant("Project Departments")) And (NewValue = vOldValue)) Then Exit Function

should help.

*** Update - you may also see this same issue with the Notes field.

December Cumulative Update Announced for Project and Project Server 2010 and 2007

Brian Smith - MSFT — Wed, 14 Dec 2011 16:20:38 GMT

Take a look over at http://blogs.technet.com/b/projectadministration/archive/2011/12/14/microsoft-project-server-and-sharepoint-server-2007-and-2010-december-2011-cu-announcement.aspx for the full details. Also don’t forget the webcast in January giving you the inside scoop on what we are delivering in these Cumulative Updates.

TechNet Webcast: Information about Microsoft Project and Project Server December 2011 Software Update (Level 200)

https://msevents.microsoft.com/CUI/EventDetail.aspx?EventID=1032493964&culture=en-us

A couple of my favorite fixes – the correction of an earlier problem that saw many non-English sites having English headers on the Project Center, My Tasks and Resource Center pages, and a change to single entry mode behavior so that you do not see unexpected changes to your timesheet entered data. More about these and all the other fixes on January 10th Webcast with Adrian Jenkins and me.

Project Server 2007: Problems with ActiveX Controls after loading the October 2011 CU

Brian Smith - MSFT — Wed, 16 Nov 2011 17:21:01 GMT

Hi All, it appears we have some issues around the download of ActiveX controls for the Project Center and other pages in Project Web Access for Project Server 2007 after loading the October 2011 Cumulative Update. Thanks to Alex for his work on this one and especially to Marc for for quickly coming up with a temporary fix. We are already working on an update to resolve this issue – but for now Marc’s fix will help you overcome the problem.

The symptom will be that you are given a message to download the ActiveX controls, which can be perfectly normal if we have updated them in a CU – however in this case the download will not complete and all the usual workarounds of manually removing and deploying etc. do not resolve the problem. The cause is that we did not deploy new ActiveX controls with this release – but we updated a version check as if we had done. So Marc’s temporary fix is just to ignore the result of the version comparison. If you happen to be upgrading from a much older SP or CU and we have since updated the ActiveX then this may give unexpected results – but certainly if you are already at a recent CU this is a good temporary solution. And of course – if you are not seeing this issue then you don’t need to make this change.

Once we have more information I will update this posting – particular around the behavior when we do fix the problem and whether you would need to revert the change first (as otherwise we may not replace the ‘updated’ file).

On to the workaround:

Navigate to C:\Program Files\Common Files\Microsoft Shared\web server extensions\12\TEMPLATE\LAYOUTS\PWA\LIBRARY
Make a safe copy of the file SHELL.JS (you may need to put this back in place before future updates)
Open the file SHELL.JS with NOTEPAD and modify the following lines:

function PJ_CompareBuildVersion(clientBuildVersion, serverBuildVersion)
{
var rgClientBuildVersion = clientBuildVersion.split(",");
var rgServerBuildVersion = serverBuildVersion.split(",");
for(var i = 0; i < rgClientBuildVersion.length-1; i++)
{
if(rgClientBuildVersion[i] > rgServerBuildVersion[i])
return 0;
if(rgClientBuildVersion[i] < rgServerBuildVersion[i])
return 0;
}
return 0

Save the file SHELL.JS
In Internet Explorer reload any of the PWA pages with the ActiveX grid and all should be good!

Sorry for any inconvenience this has caused – and don’t hesitate to contact Microsoft if you have any further issues with this cumulative update.

Q and A following the recent October Cumulative Update Webcast

Brian Smith - MSFT — Mon, 14 Nov 2011 21:34:00 GMT

(Managed to hit a key combination in Windows Live Writer that posted this half way through – sorry if this is a duplicate…)

Thanks to everyone who attended the webcast covering details of the Project and Project Server 2007 and 2010 cumulative updates for October 2011 – as well as some brief comments on Project and Project Server 2007 SP3. The recorded webcast can be viewed here - https://msevents.microsoft.com/CUI/EventDetail.aspx?EventID=1032493963&Culture=en-US. Mark your calendars for the next one about the December 2011 CU on 1/10/2012 - https://msevents.microsoft.com/CUI/EventDetail.aspx?EventID=1032493964&culture=en-us.

I’m posting the questions that were asked both during and at the end of the webcast – in some cases I’ve found out more complete answers since the webcast and have included the detail here too.

Q.   Adrian mentioned an anti-virus package had slowed down Project Professional – which one?

A.   We not going to name names here – and really the same could apply to any of the packages that scan what winproj is doing. It is really up to the customers to decide if they feel they can exclude the project executable from the virus scanner – balancing the virus risk against the performance gain.

Q.   Ref fix 30990 (Modifying time-phased work in Project Professional moves actual work; additionally, the "Out of Sync" message does not occur), will this fix work if you are on a server with August 2011CU+SP1, and client with SP1 + October 2011CU? (i.e. out of sync client server CUs)

A.   Yes, this is a client side fix and not dependent on the server CU or SP level.

Q.   Project Server 2010...Graphical indicators - we have had lots of issues with graphical indicators not displaying correctly. The issue is random (as far as we can tell) and it doesn't matter whether the custom field references a formula or lookup table. Both can be affected. Sometimes (not always) opening, modifying, saving and publishing the project in PWA resolves the issue, but on next publishing from ProjectPro the indicators disappear again. This problem is well over 18 months old and was supposed to be fixed in the Dec 2010 update, but wasn't. When will this be fixed ??? Customers (as well as ourselves) are finding this incredibly frustrating.

A.   Had some dialog on this after the webcast with a couple of the attendees (Thanks Kelvin and Marc!). Looks like we have specific issues where the first publish does not show the indicator – just the text. We have a fix coming in December that should address the publish from Pro causes the loss of the GI – I’ll keep you posted on other changes…

Q. Is it true that after April there will be no additional CU's for 2007? Followed by - My understanding is that Primary Support for Project Server 2007 ends in April 2012

A.   At the time of the webcast the Microsoft Support site at http://support.microsoft.com/lifecycle/?p1=11381 listed April 2012 as the date for the end of Mainstream Support. This has now been updated to reflect the recent SP3 release – so that makes the end of mainstream support date 10/9/2012. This would tend to indicate that the August 2012 CU might be the last for 2007, though if hotfix requests are accepted right up to the end of mainstream support I could see there also being an October one. As this is the first CU supported version of Project to go out of mainstream support this is new territory. As you will have noticed from this webcast – the number of fixes for 2007 has dropped off quite dramatically.

Q. Project Demo Image–Excel Services problem after SP1 upgrade http://epmsource.com/2011/07/16/project-demo-imageexcel-services-problem-after-sp1-upgrade/ has this been fixed?

A. This is fixed for the latest partner-only demo – see http://blogs.msdn.com/b/jkalis/archive/2011/10/26/project-2010-demo-pack-version-2-now-available-to-partners.aspx

Also we received a suggestion that we should include any SharePoint updates that would have an affect on PWA or project sites. We will see if we can build this in next time.

And finally – the PowerPoint didn’t get loaded up to the Live Meeting – so I am including here.

Microsoft Project Server and SharePoint 2007 and 2010 October CU 2011 over on the Admin blog

Brian Smith - MSFT — Wed, 02 Nov 2011 16:44:11 GMT

The Cumulative updates for October 2011 for both the 2007 and 2010 versions have now been released. Full details over on the admin blog at http://blogs.technet.com/b/projectadministration/archive/2011/11/02/microsoft-project-server-and-sharepoint-2007-and-2010-october-cu-2011-announcement.aspx

I’ve also included the usual details on the 2007 CU which was released last week (The packages containing SharePoint Foundation were withdrawn and re-released earlier this week).

Enjoy!

Project 2007: Service Pack 3 (SP3) and the October 2011 Cumulative Update (CU)

Brian Smith - MSFT — Fri, 28 Oct 2011 16:01:00 GMT

*** Update - thanks to all who corrected me on the title - don't know how many times I must have read that and not noticed - yes, this is for 2007 and not 2003... ***

I’ve posted over on the admin blog - http://blogs.technet.com/b/projectadministration/archive/2011/10/28/project-server-2007-service-pack-3-and-the-october-2011-cu-announced.aspx so go there for more details.

*** Update – just heard there are a couple of issues being investigated with the 2007 Office packages – this affects the MOSS Server package 2596540 (which is the roll up that also includes the Project Server 2007 fix below), and a Project 2007 fix that was specifically aimed at implementations of Project 2007 and Office 2010 where Visual Reports would not work. Also corrected the description below from 2010 to 2007 – sorry… Both of the updates below are unaffected. *** All affected fixes have been re-released 11/1/2011

There is a slight delay on the release of the 2010 October 2011 Cumulative Updates – more details when I have them.

Don’t miss the next Cumulative Update webcast where we (or more correctly, Adrian) will also talk about SP3 (briefly – as there isn’t much to say)

Here’s the URL for the 11/8/2011 8:00:00 AM - Information about Microsoft Project and Project Server October 2011 Software Update
https://msevents.microsoft.com/CUI/EventDetail.aspx?EventID=1032493962&Culture=en-US

Project Server 2010: August Cumulative Update package hotfix re-released

Brian Smith - MSFT — Tue, 18 Oct 2011 19:24:16 GMT

Pink Floyd are not the only ones re-releasing their back catalog this month – we have re-released the SharePoint Foundation 2010 August Cumulative Update. As this is also contained in the roll-up packages then these too have been re-released. Unfortunately the title of the Project Server package has also been updated to reflect this new release date and it could make it look like the October Cumulative update released early – so take care!

The re-released Project Server package at http://support.microsoft.com/kb/2553049 is still the August Cumulative Update, although it is labeled Description of the Project Server 2010 cumulative update package (Project server-package): October 13, 2011. This is NOT the October CU coming early! The version has been incremented to 14.0.6109.5005 to reflect the new build, but it is just the SharePoint Foundation component that has been updated. The Project Server only package at http://support.microsoft.com/kb/2553047 is still showing the original date and version, as it was not rebuilt as it does not contain the SharePoint Foundation component. (sts-x-none.msp).

I think the change in the SharePoint package is support for updates to Russia, Fiji and Samoa daylight saving time (DST) settings in SharePoint Foundation 2010 – I’ll see if I can get confirmation.

Project Server 2010: Don’t Trash the Cache!

Brian Smith - MSFT — Tue, 11 Oct 2011 00:16:00 GMT

I was going to use the title asking ‘why are people still deleting the cache?’ until my colleague Corrie came up with this much better one! Rather than asking why you are still doing it – I am telling you not to!

I know there is a lot of history behind this one, and for those of you that used Project Server 2007 in its early days there were some challenges such as the ‘check-in pending’ saga that got people in to the habit of deleting the project cache. We fixed the problem, then we fixed it again (and again) and you should not generally be seeing any issues with leaving your cache alone to do its job. However, many customers I talk to are routinely deleting the project from the local cache before they open it and then again after they close it! Why!?! Its job is an important one – it saves you having to pull that data from the server again – which will reduce network traffic, the hit on both the web services and the database, which means they can be getting on and doing useful stuff.

I’ll also address a miss-conception misconception (thank you Trevor - not addressing the misconception that I can spell...) here that I have heard from a number of customers – the choice of where to load the project from – cache or server? You don’t have a choice – Project will load it from the cache if it is there, and then load any incremental pieces it needs from the server, to get you the current version of that plan. In the screen shot below:

the line actually reads ‘Retrieve the list of all projects from Project Server’. It does not also read – ‘…and open any I might choose after clicking this link from the server and ignore the local cache’. You don’t get the choice and you don’t need to choose.

I’m sure many of you will not have read this far before clicking the comments option to tell me of all the problems you are having.(and I’m sure some of you are still having problems). First check that you have the latest cumulative updates and service packs. If you are still really having issues unless you delete the local cached copy then we certainly need to hear about it so we can fix the problem rather than have you waste your time and system resources doing things that you should not need to do.

I will admit that there can be times when as support engineers we will ask you to remove your local cache to troubleshoot specific scenarios. The cache itself also has intelligence that allows it to decide that it may have some bad stuff – and it will get a new clean copy of data from the server (symptom of this will be several files in the cache directory with 1,2,3 etc. at the end). There have also been a few bugs we have worked on recently which ONLY surface when the user has cleared their cache!

So please, if you have been deleting your cache as a matter of routine, then either stop – or speak to your PMO or IT people and ask why they have you do this – and if we need to fix something else then we can take a look.

Project Server: Post SP1 Cumulative Update Webcast Series

Brian Smith - MSFT — Tue, 27 Sep 2011 13:34:23 GMT

In a little over an hour, at 8:00 AM PST, the first of the post SP1 webcasts will start – presented by Adrian Jenkins and me. The SP1 webcast delivered in July also included coverage of the June Cumulative Update. We will be talking about both Project Professional and Project Server, and the 2007 and 2010 releases.

Here is the link for the August 2011 Cumulative Update TechNet webcast titled Information about Microsoft Project and Project Server August 2011 Software Update - https://msevents.microsoft.com/CUI/EventDetail.aspx?EventID=1032493921&Culture=en-US

And the rest of the series can be found at the following links.

Here’s the URL for the 11/8/2011 8:00:00 AM - Information about Microsoft Project and Project Server October 2011 Software Update

https://msevents.microsoft.com/CUI/EventDetail.aspx?EventID=1032493962&Culture=en-US

Here’s the URL for the 1/10/2012 8:00:00 AM - Information about Microsoft Project and Project Server December 2011 Software Update

https://msevents.microsoft.com/CUI/EventDetail.aspx?EventID=1032493964&Culture=en-US

Here’s the URL for the 3/13/2012 8:00:00 AM - Information about Microsoft Project and Project Server February 2012 Software Update

https://msevents.microsoft.com/CUI/EventDetail.aspx?EventID=1032493966&Culture=en-US

Here’s the URL for the 5/8/2012 8:00:00 AM - Information about Microsoft Project and Project Server April 2012 Software Update

https://msevents.microsoft.com/CUI/EventDetail.aspx?EventID=1032493968&Culture=en-US

Join us if you can – or listen to the recording if you can’t make it – available later on the same URL.

Project Server 2010: Installing Project Server to an existing SharePoint Server farm

Brian Smith - MSFT — Tue, 13 Sep 2011 23:20:31 GMT

I know that unless you load all the bits before running the Configuration Wizard then you will always add Project Server to an existing SharePoint Server farm – but in this blog I am specifically talking about adding Project Server to a farm that has been up and running for a while – and potentially has Service Packs and Cumulative updates loaded.

The best way to do this is to create a slipstream installation – and there are plenty of resources out there that explain this process – but basically you extract your various updates and put them in the Updates folder under your install source and these will get applied as the installation proceeds. However, in researching this topic I learned that this isn’t the only way so thought it was worth sharing. You can load the original release version of Project Server (RTM – release to manufacture) even if your farm is at the SP1 level plus cumulative updates. In reality I only tested to August CU, and I am sure there will come a point where this will not be practical (SP2 would block an RTM install) – but for now it certainly works and would be supported. That said – it would be good practice to bring the farm up to a level where all the components were at the same release level. For information on the release level of each component you can go to Central Administration, Upgrade and Migration and Check Product and Patch Installation Status. I have a ton of language packs loaded so I won’t give you a full screen shot(s) but the foot of mine looks like this:

So you can see the version installed for each component (usually matching the RTM or last Service Pack, as in this case) as well as the Cumulative Updates (June superseded by August) along with useful links to the KB articles. In my case I do have Project already loaded and updated to August CU.

At the top of the page there is also a link to the latest updates - Click here for the latest information on available updates for SharePoint 2010 Products

Going to that page also has the link for Project updates – which is Updates for Project Server 2010. You will also see on this screenshot that the Galician/Galego language pack is still at the original release version of 14.0.4763.1028. This language pack (and Basque) does not yet have the language pack service pack released – but it is still possible to update to SP1 for the farm without having to update all the language packs.

That was just an aside on versions and where to find them – but the main point is that you can load RTM Project Server to a farm and use Project Server. It will get loaded at the original release version even if you had already loaded the rollup Service Pack (SP) and/or Cumulative Updates (CU) that included Project Server – the Project files will not have been applied. Once you are ready to load the SP/CU you will need to re-load it and run the configuration wizard. You will not get any warnings or errors even though you feel are re-installing something you already applied – it correctly recognizes that there is new stuff to update. The only slight exception are the language pack service packs – which you do not need to reload – and Project Server will benefit from any language packs (and language pack service packs) loaded before it was installed – so no need to re-install those either.

Another good thing to check is the database status – once you have run the config wizard all should be good – but if you forget then you will see messages Database is too old and upgrade is required as mentioned in my previous blog posting - Project Server 2010- Can I delay running the SharePoint Configuration Wizard-.

Project Server 2010: Are your prints from PWA blank? Try turning off IE compatibility.

Brian Smith - MSFT — Thu, 08 Sep 2011 20:46:00 GMT

Quick posting today – we had a customer seeing issues when printing from the Project Web App schedule web part where if they went over a certain number of rows (around 42 – spooky!) the printed page was blank (although the pop-up with just the grid displayed for printing looked just fine). This was after seeing this warning – which is expected – when you are printing a grid of more than 30 rows: Print Warning - There are more than 30 records in the current grid. Preparing the print page for this many records may cause the browser to alert that the page is running slowly.

In my testing internally I saw the same, I clicked OK on the warning, the new page opened with everything looking good and I could choose my printer – but the print was blank (I did try a large plan and in this case the final 40 or so tasks printed…)

However, if I turned off the compatibility setting under Tools, Compatibility View Settings, so that these were not applied to Intranet sites then all worked as expected and I got a good print. If you have explicitly set your PWA sites in compatibility you may see the same issue. If you really need prints though you may find Project Professional more flexible in giving you a good printed page.

*** Update 9/12/2011 I should point out this is based on IE 8/9 - it does appear that IE 7 has some issues with the PWA printing that changing settings cannot fix ***.

Project Server: SharePoint Installation issues when using FIPS

Brian Smith - MSFT — Fri, 02 Sep 2011 21:40:23 GMT

Most days I learn something new, and yesterday was no exception. I was working with one of our Senior Consultants, Rob Bowers, on an installation problem. The SharePoint Configuration Wizard was failing during the initial configuration of the farm on step 3. The error that came up was:

Configuration Failed. One or more configuration settings failed. Completed configuration settings will not be rolled back. Resolve the problem and run this configuration wizard again. The following contains detailed information about the failure:

Failed to create the configuration database.

An exception of type System.InvalidOperationException was thrown. Additional exception information: This implementation is not part of the Windows Platform FIPS validated cryptographic algorithms.

In the PSCDiagnostics log created during the execution of the wizard the same errors could be seen – the first was:

Task configdb has failed with an unknown exception , followed by

Exception: System.InvalidOperationException: This implementation is not part of the Windows Platform FIPS validated cryptographic algorithms.
   at System.Security.Cryptography.SHA256Managed..ctor()
   at Microsoft.SharePoint.UserCode.SPSolutionValidatorCollection.ComputeHash()
   at Microsoft.SharePoint.Administration.SPUserCodeService.UpdateValidatorsHash()
   at Microsoft.SharePoint.Administration.SPPersistedChildCollection`1.Add(T newObj, Boolean ensure)…

A quick search (Bing of course) found that FIPS was referring to the Federal Information Processing Standard (FIPS) 140-2, Security Requirements for Cryptographic Modules. An article on TechNet http://technet.microsoft.com/en-us/library/cc263215.aspx has a security note that mentions some potential issues with workflows – but not the failure in the configuration wizard. Another great link from Mahesh Srinivasan at http://collaborationsharepoint.blogspot.com/2010/10/this-implementation-is-not-part-of.html helped move things in the right direction. Even with FIPS not enabled through group policy settings there can still be registry keys set that are enabling some of the features. In Rob’s case, like Mahesh, he found that the two keys were set to 1 – enabled, and a third key was set to 0 – disabled. The keys were:

HKLM\SYSTEM\ControlSet001\Control\LSA\FipsAlgorithm
HKLM\SYSTEM\ControlSet002\Control\LSA\FipsAlgorithm
HKLM\SYSTEM\CurrentControlSet\Control\LSA\FipsAlgorithm

Until each of these was set to 0 the error above blocked running of the configuration wizard. Remember, any time you are changing registry keys you should take back-ups. Obviously this change is something you should to talk to your platform security team about too – as if you are changing these values you may need to get an exception to your company’s hardening policy for your SharePoint servers.

FIBS 104-2 is intended to ensure that only only validated cryptographic modules are used in software when securing data. SharePoint uses cryptographic modules, for example MD5, that are not validated – but it is in fact not using them to secure data but to create hash values that are used as unique identifiers. It is this action that FIPS is blocking that causes the failure in the configuration wizard.

For more information on FIPS 104-2 see http://csrc.nist.gov/publications/fips/fips140-2/fips1402.pdf, and for general FIPS information see http://csrc.nist.gov/publications/PubsFIPS.html.

Project Server: Why do my date selections lose a day?

Brian Smith - MSFT — Fri, 02 Sep 2011 16:51:00 GMT

*** Update - should probably have mentioned this more prominently - but this DOES NOT affect schedule dates - just the display date ranges in Resource Plans. When I re-read it after posting I realised this might look quite alarming***

This was an interesting issue that Adrian Jenkins finally cracked – and I had seen postings on the forums and we had some internal customer cases. It appeared to be related to time zones as the problem had only occurred in Europe and Eastward. So if your time zone is left of Greenwich (Sorry – UTC) you probably don’t need to read any further. But as this may work in reverse also – might just be worth seeing why this happened.

In my example below, I have my server and SharePoint farm running at UTC+2 – Athens, Bucharest and Istanbul (and others of course). I’ve also set the regional settings to Greek (but don’t have the language pack) – just so the date format is the same format as most people this affects are using. If you are in the Resource Plan and want to change the date range and say you select 1/12/2011 to 31/12/2011:

you will find when clicking OK that the actual dates are set to 30/11/2011 and 30/12/2011

Looking behind the scenes we can see that the actual date (and time!) saved is midnight. This is persisted in the User Properties table – as it isn’t setting the resource plan start and end date – just the current window of dates that you wish to review. Worth remembering that there can be stuff before and after these dates… but back to the plot. These dates are stored in a numeric string representing the number of milliseconds since 1/1/1970. Between selecting, saving and recalling there is a conversion according to the current time zone and basically it thinks that at midnight on whatever day you choose (assuming you are UTC+X) – it was still yesterday at UTC! In reality the time stored is the UTC time taking in to account the time zone difference – so in my example I set 1/12/2011 (it assumes time of 0:00) but what was saved was actually 2 hours earlier 30/11/2011 22:00.

Easy workarounds – either put in the day after you really want – or add a time value to the string. So I could enter:

and this would then give me the right dates.

This shouldn’t affect anyone West of Greenwich – however, if I put in a time here in the Pacific Time Zone that was after midnight UTC – such as 5pm, then my date would go forward to the next day!

Sorry for any confusion this has caused. I should also point out that although I am specifically talking about the date range for the resource plan view this could also affect other pages too – but generally only dates persisted as part of a view. Rest assured that in my testing it does not affect dates entered in the schedule grid.

Project Server: August CU Released for 2007 and 2010

Brian Smith - MSFT — Thu, 01 Sep 2011 15:45:00 GMT

Full story over on Microsoft Project Server and SharePoint 2007 and 2010 August CU 2011 Announcement and a slight delay on the server rollup packages for 2010 (a week or so), but the Project and Project Server 2010 package is there if you are waiting on a fix and want to get started on your testing. *** Update - 9/7/2011 - now posted *** Don’t forget SP1 too! More details of SP1 can be found at http://blogs.msdn.com/b/project/archive/2011/06/28/announcing-the-release-of-service-pack-1-sp1-for-microsoft-project-and-project-server-2010.aspx and the downloads listed at http://support.microsoft.com/kb/2510766. Q&A from the SP1 TechNet webcast can be found at http://blogs.msdn.com/b/brismith/archive/2011/07/11/project-server-2010-sp1-and-june-cu-questions-and-answers.aspx.

Project Server 2010: Timesheets - Where do Personal Tasks come from?

Brian Smith - MSFT — Wed, 31 Aug 2011 18:41:59 GMT

If you have the setting to allow personal tasks to be added to the timesheet then they of course could come from the user. But what about the ones that haven’t been entered by the user and appear to have the same name as tasks that were in projects that they had already started work on? We have had a few questions on this behavior so thought it worth a blog posting.

So here is my timesheet – and I have 5 tasks in the “Project BriSmith TS Blog” project, and I’ve spent an hour on each – so I’ve entered the time and saved my timesheet. (Status is “In Progress”)

Later on I go back to my timesheet and things have changed!

So what happened? The task names give you the clue, and the Project Manager (me) has followed through on his action for each of the tasks. Basically if a task gets deleted while a timesheet is in progress then we do not delete the time that is already entered but instead the task gets changed to a Personal Task. A couple of interesting things to note though – the task name is actually cached as part of the timesheet (in case it gets deleted!) so the spelling mistake was actually corrected in the plan for the “Correcttting the spellling” task – but the cached version is still shown. This would be the case for any re-named task – as long as the resource was still assigned. If I click through to the task I see the current name published from the plan – correcting the spelling:

The next thing to note is that if I delete and re-create a task then as far as project is concerned this is a new task (with a new unique GUID) so it shows up as a Personal Task (keeping the original entered time) and also as the new task.

The 4th task is interesting – I unassigned myself from this task and then re-assigned – so this is actually a new assignment – but as far as the timesheet is concerned it still shows it related to the project. Also if I had published after de-assigning I would have seen a Personal Task and then after re-assigning and publishing this personal task would be gone and the project task back again (along with the entered time). This surprised me a little - so I had to take a look behind the scenes. It appears that as long as you don’t assign this task to anyone else, then you will get the same GUID for the re-assignment – but if you re-assigned to someone else, and then added back the original resource then the GUID would change and the timesheet would show the same behavior as the deleted and re-created task – both a Personal Task with the already entered time and the ‘new’ project task.

The 5th task just got deleted so no longer exists so my recorded time now shows against Personal Tasks. And finally the untouched task is still there as expected.

So if the Personal Task does really relate to a task that has been re-created in some way then the resource should copy the time entered across the the project task and delete the personal task (Remove Task in the ribbon). If it has been removed for good then the personal task is a record of the time spent – and should remain (depending on your time recording processes).

This is the behavior when the timesheet is “In Progress”. If it is approved then the Project name and Task name are persisted even if the task is deleted.

Other actions can lead to this behavior – if the project is saved back to the server (save as) over the top of an existing plan then all tasks and assignments will be replaced by identical ones – and any tasks in progress will change to Personal Tasks in the timesheets. Also before we fixed the bug with Save for Sharing in the June CU this would change the GUIDs too – and have the same effect.

All the above behavior assumes that Single Entry Mode is turned off. With Single Entry Mode turned on then the Personal Tasks do not appear – the tasks just go (along with the time entered into them) – and also any task name changes reflect immediately in the timesheet (it is a direct link to the task, rather than a placeholder in the timesheet). The re-assignment to the same task, even though it maintains the same GUID does not keep the entered time. Single Entry Mode keeps the My Tasks and Timesheet view in Sync and Personal Tasks are never seen on the My Tasks view.

Project Server 2010: SharePoint Permissions in upgraded PWA instances

Brian Smith - MSFT — Tue, 23 Aug 2011 17:42:10 GMT

I was reviewing a case yesterday and just checking the behavior of SharePoint permissions in Project Web App after an upgrade from 2007 to 2010 – either in place or a 5 database – thought this might be useful information to blog about.

In 2007 in the Project Web Access site the users would be added directly to the site with SharePoint permissions based on their roles. These permission levels were identifiable by their names which would be like Project Managers (Microsoft Office Project Server). In the Project workspaces these same permission levels would be used, and the users again added directly based on their roles within the project. So this generally be a small subset of the overall user population – just those added as resources on the plan, and potentially others depending on the use of the View Project Workspace permission within categories. Here is the start of the list of permissions on the PWA site – Project workspaces will be similar although the permission levels may be different.

In 2010 in Project Web App we now use groups within the /PWA site and the users are added in to these groups rather than directly to the site. The groups are named Project Managers Group (Microsoft Project Server), Team members group (Microsoft Project Server) etc. Some individual user accounts will be present in PWA as shown below (blanked out) – these will be site collection administrators and other farm admins.

For the Project sites (formerly called workspaces) we still add the users directly, but we have new SharePoint permission levels which are named similarly to 2007, but we have dropped ‘Office’ from the product name – so they are now like Readers (Microsoft Project Server), Web Administrators (Microsoft Project Server).

If you upgraded from 2007 to 2010, either in place or 5 DB then you will retain the old permission structure – but also get the new ones too. So for example you would see all your users individually in the Project Web App permissions, as well as them belonging to the respective groups. At the Project Site level you would see the users but with both Permission Levels applied – so they might have Web Administrators )Microsoft Office Project Server), Web Administrators (Microsoft Project Server).

All of this is the expected behavior and none of this will break anything, although I admit it may look unusual having the two sets of permission levels. Also it does not leave any security issues as in 2010 if I remove someone from a Project plan then they get taken off the site with both sets of permission levels – and if I inactivate a user they are removed both at the individual level from PWA and also removed from the group that 2010 would have put them in – so all is good. If you really wanted to then you could delete the permission levels with (Microsoft Office Project Server) in the names, and also remove the individual user permissions on PWA (apart from the site collection admins of course). Worth checking that the groups contain the members you expect first just in case for some reason the sync hasn’t populated the groups yet.

Project Server 2010: Can I delay running the SharePoint Configuration Wizard?

Brian Smith - MSFT — Tue, 23 Aug 2011 17:04:00 GMT

In SharePoint 2010 when you upgrade to a new Cumulative Update or Service Pack this involves two main steps – first is loading the binaries, and the second is running the SharePoint Server 2010 configuration wizard (psconfig or psconfigui). SharePoint supports delaying the running of the configuration wizard or even detaching content databases while running the wizard, so that these perhaps slow parts of the process can be managed in a more timely fashion. If you look at a SharePoint farm that is in this condition you will see in Central Administration, Upgrade and Migration, Review database status, that it says against many of the databases (Content, Config, and Admin Content) – Database is in compatibility range and upgrade is recommended. However, if you have Project Server installed then you will see against all of its databases (certainly for SP1/June CU) – Database is too old and upgrade is required. Some other databases such as BDC or PPS ones may just say No action is required if there were no updates for schema in the particular release. For some CUs you might see this for Project and the SharePoint content databases too.

If you ignore this message and try and go to PWA then you will get an error message: Error, Project Web App cannot connect to Project Server. For more information, contact your system administrator. – along with a GUID for tracking the full error in the logs.

Looking in the logs you will find the following exception and unexpected level records – which are pretty self explanatory.

08/23/2011 09:46:41.85 w3wp.exe (0x1724) 0x0FC0 Project Server General g7ls Exception System.ServiceModel.FaultException`1[Microsoft.Office.Project.Server.Interfaces.DefaultServerFault]: The databases are out of the range of compatibility, upgrade your databases. (Fault Detail is equal to Microsoft.Office.Project.Server.Interfaces.DefaultServerFault). fe5f9380-1f54-4021-a6a2-5fe7d3e321e8

08/23/2011 09:46:41.85 w3wp.exe (0x1724) 0x0FC0 SharePoint Foundation Runtime tkau Unexpected System.ServiceModel.FaultException`1[[Microsoft.Office.Project.Server.Interfaces.DefaultServerFault, Microsoft.Office.Project.Server.Communications, Version=14.0.0.0, Culture=neutral, PublicKeyToken=71e9bce111e9429c]]: The databases are out of the range of compatibility, upgrade your databases. Server stack trace: at System.ServiceModel.Channels.ServiceChannel.HandleReply(ProxyOperationRuntime operation, ProxyRpc& rpc) …

So to answer the question in the title – No, you cannot delay the running of the configuration wizard if you are using Project Server if there are database updates required in the particular patch you have loaded. Not every CU will require database changes – but rememeber these are cumulative, so the need for the database update will also depend at what level your server is when applying the patch.

*** Update based on feedback from one of my favorite customers - if you really cannot run the config wizard for a while then it will only be PWA that won't work. The Project sites will still work (apart perhaps from issues/risks/deliverables lists) and of course all the other SharePoint features and Central Administration will be just fine. ***

Microsoft Project 2010: Cumulative Update Version Blog Updated

Brian Smith - MSFT — Mon, 22 Aug 2011 16:12:05 GMT

Just for reference – I have added the April and June CU information, and some notes around SP1 versions, to the on-going blog post at http://blogs.msdn.com/b/brismith/archive/2010/09/23/how-to-tell-which-cumulative-update-hotfix-or-service-pack-version-of-project-server-2010-and-project-2010-you-are-running.aspx.

Microsoft Project Server and SharePoint 2007 and 2010 June CU 2011 announcement over on the Project Administration Blog

Brian Smith - MSFT — Mon, 22 Aug 2011 15:31:35 GMT

Better late than never. We realized (with a nudge from Treb) that we hadn’t posted the normal announcement – so now we have – over at Microsoft Project Server and SharePoint 2007 and 2010 June CU 2011 announcement.

Project Server 2010: Portfolio Analyses–How are equal priority projects chosen for resourcing?

Brian Smith - MSFT — Fri, 19 Aug 2011 22:59:39 GMT

A bit of fun for Friday afternoon. A question came via the forums on the subject of Portfolio Analysis in Project Server 2010 and how a decision would be made on the allocation of resources if the projects had exactly the same priority. Firstly there is no real optimization logic in the resourcing part of the analysis – most of the heavy lifting happens in the cost analysis – where other metrics can also be taken into account. In terms of the resourcing this just happens in order of priority – so the top project gets resourced, then the next, and so on. For any project where there is insufficient of any type of resource then that gets forced out. But what if you have equal priority projects? In truth the Portfolio Manager would probably make the call and force in or out appropriately – but it is still technically interesting to know how it happens (well I wanted to know anyway…)

I had a play around with this today. So in my scenario I was also interested in seeing how equal cost choices were differentiated as well as equal priority ones chosen to get resourced. And the interesting result is that each appears to make the choice in the totally opposite way!

If I have 4 projects of equal cost and equal priority then the order of selection if the budget is not available to do all of them is carried out in the order of their GUIDs (PROJ_UID from the SQL Server database table MSP_PROJECT) - but not in the strict order that SQL might use - that sorts by the lowest order grouping of the last 6 bytes, but in the more straightforward left to right order of the hexadecimal characters.

As an example - SQL Server would put D741FE6E-D426-4A41-8BCC-370FDE23A3E4 before 6EDBF314-BFAE-4838-8CBC-6B95E91EC0C5 based on the last group (37.. before 6B..) but the order the Portfolio Analysis uses would take 6E... before D7... So almost random, unless you happen to know (and care) what your GUIDs are.

Now on to the resourcing. This happens in entirely the opposite order. So the 'highest' value of GUID (still using the full left to right logic) will get resourced first and the lowest gets resourced last. Again - it is really irrelevant how this happens and I'd guess you'd be forcing things in and out based on other criteria rather than some random 32 character hexadecimal string.

If you are keeping up - or slightly ahead of me - this opens up an interesting paradox. If I have 4 projects that have equal cost ($20,000 each) and equal priority - and I have a budget of $80,000 - but they each require a resource that I only have one of - then the project with the highest GUID will get selected for resourcing! However, if my budget is just $79,000 then that very same project would be the one that was rejected at the cost analysis stage - leaving the next highest to get resourced!

The Mystery Behind SharePoint 2010 Patching

Brian Smith - MSFT — Fri, 12 Aug 2011 14:49:37 GMT

I came across this great posting from Jie Li while looking around Stefan Goßner’s blog at http://blogs.technet.com/b/stefan_gossner/ and thought it would make interesting reading for the Project audience too. One word of caution first – Jie Li mentions the well known SharePoint practice of detaching content databases while applying patches so that you can upgrade them at your leisure later – once you have re-attached them. Project Server databases (Archive, Draft, Published and Reporting) ARE NOT content databases – please do no detach these while patching or you will get into an inconsistent state that may cause you (and then me and my support colleagues) big problems…

On to the article - http://blogs.msdn.com/b/opal/archive/2011/06/30/the-mystery-behind-sharepoint-2010-patching.aspx.

Enjoy!

Project Server 2010: How to best manage large numbers of resources

Brian Smith - MSFT — Wed, 10 Aug 2011 23:29:00 GMT

This posting follows on from the one yesterday concerning how Project Server makes use of SharePoint permissions and features – but concentrates on some potential issues you can run into if you have a very large user base and also have projects that have very large teams. We are also authoring a TechNet article explaining this in more depth – I will add a link once it is published. This isn’t going in to the usage of the RBS or the other internal feature – but concentrates more on the technical issues of large user populations. If you fit in this category then read on…

As mentioned in the previous post Project Server 2010 uses the normal SharePoint permissions infrastructure to set access control both to the Project Web App (PWA) site and also any Project sites that are created for the individual project plans held in Project Server. At the PWA site level the users are added to certain groups depending on their permissions levels within Project Server, so you will generally see SharePoint groups for Project Managers, Readers, Team members, Web Administrators and finally Workflow and Project Detail Pages Administrators. Each of these groups will show the individual PWA users as appropriate. This is a change from 2007 where individuals were added to the PWA site with specific permission levels. You may have seen issues in 2007 if you had large numbers of users as whenever changes were needed in the member permissions the users would be removed and then added back – so some users would get “Access Denied” until they were added back after a change. We had some workarounds for this scenario involving turning off the user synchronization. In 2010 we made a couple of changes to avoid this problem – firstly the change to using groups at the PWA site level, and secondly we now remove then add back each individual as opposed to removing everyone and then adding back everyone. So getting an Access Denied in the same scenario in 2010 is very unlikely.

At the Project site level however we do not use the group approach and manage the users on an individual basis. In most scenarios this is not an issue as the number of resources assigned to a project, and hence added to a site, is generally low compared to the total number of users in the system. However there could be some scenarios where customers wish to have many or all of their users accessing many or all of their project sites. This could either be achieved by adding many users to a project – or by giving the “View Project Site” permission at the team member level in a category that included many or all projects. Either way this would then add very many individual users with permissions to the project sites. And why is this a problem? If the numbers of users is large then it is possible for the recommended software boundaries and limits of SharePoint Server to be exceeded – and this can lead to performance issues. Each user added individually to a site would be considered a security scope – and the recommended maximum number of unique security scopes per list is 1,000 (SharePoint Server 2010 capacity management – Software boundaries and limits - http://technet.microsoft.com/en-us/library/cc262787.aspx). So each list and library in the site would be inheriting from the parent site permissions – and would exceed this limit if more than 1,000 user had access to the site (as they are individually added).

In our experience the performance issues would then relate to any change in the site membership caused by changes in the categories or groups – or following such actions as adding a user or inactivating a user. For example this last action of inactivating a user will actually remove that user from all sites they have access to – and the reason the limit is imposed is that when it is exceeded the process of removing a user can become very slow – particularly if this same user is also being removed from very many sites each of which is also way over the limit. In extreme cases with multiple user inactivations it is possible that the server will become unresponsive and unable to authenticate users. I will include some of the error messages you might see, and the corresponding ULS entries at the end of this posting so that this aids finding this potential cause.

If you are following along (and I’m sure some of my readers are way ahead of me…) you will realize there is a Catch-22 here. Your server could become unresponsive whenever you need to manage users because you have too many users with permissions on the sites. So remove some users… which will then make the server unresponsive... How to escape from this loop? There are some quick ways to get this resolved – but before rushing in to that it is better to review what it is you are really trying to achieve. If the desire is that most people can access most projects then managing the permissions outside of Project Server using groups and inheritance from PWA is the way to go. If however the fact that many users had access to many sites was really a mistake then you need to correct that issue – and either remove the “View Project Sites” from the offending category or reduce the number of resources assigned to the plans – but first of course you need to stop the synchronization of users to the sites otherwise any action may make your server very slow. As mentioned in the previous post this can be achieved by us of the UserSyncSettings method – and just repeating that here to save you having to open that post up:

The setting can be changed using the PSI and the Admin Web Service and the UserSyncSettings method. The enumeration of values that can be set are detailed at http://msdn.microsoft.com/en-us/library/websvcadmin.usersyncsettings_di_pj14mref.aspx, and the method described at http://msdn.microsoft.com/en-us/library/gg229480.aspx.

Turning off the Project Site sync is achieved by the enumeration DisablePWS.

Member name	Description
Enabled	Value=1. Enable all synchronizations.
DisablePWA	Value=2. Disable synchronization with Project Web App.
DisablePWS	Value=4. Disable synchronization with project sites for the user.
DisableEmailSync	Value=8. Disable email synchronization.
DisableAll	Value=16. Disable all synchronizations.

*** CORRECTION *** Looks like this table may be in error both here and in the SDK. Still awaiting confirmation but looks like the values should be 0,1,2,3 and 4 - rather than 1,2,4,8,16. ***

This relates to settings in the MSP_WEB_ADMIN table of the Published database in the WADMIN_USER_SYNC_SETTING column. So for example a query such as:

Update [ProjectServer_Published].[dbo].[msp_web_admin] set [WADMIN_USER_SYNC_SETTING] =4

would do the same as using the method to set the enumeration:

int syncSettings = (int)SvcAdmin.UserSyncSettings.DisablePWS;

We’d certainly prefer you to not touch the DB correctly – but I’m guessing that many of you would find it much easier to execute a SQL query to update the value than to write the code necessary to do the same (I certainly would!).

Once that is turned off then you can safely do user management without causing further performance problems – but of course it would still be possible to trigger the same issues if you tried removing the users directly from the project site, using the out of the box SharePoint functionality. One way to remove the users very quickly without triggering the individual deletion that causes the problem is to inherit permissions from the parent. This can be done via the UI on the Site Actions, Site Permissions page of the individual sites:

This will lose any custom permissions. If your end goal is to give most users access to most sites then this may be how you want to keep things long term – so before taking this action you would probably want to be sure that the PWA site has the right permissions for all the users who need access – as that will be where this site will start inheriting from once you click the button. Obviously if you have thousands of sites (and you probably have if this is causing you problems) then PowerShell can automate the change for you.

Run the following PowerShell command in the SharePoint 2010 Management Shell

$site = Get-SPSite “<url of PWA>”

Foreach ($web in $site.AllWebs) {

        $web.Update()

        $web.ResetRoleInheritance()

        $web.Update()

        }

$site.Dispose()

This (or the UI method) would also need to be run for any new sites created to avoid the problem coming back – but if you had ‘corrected’ your categories and team memberships then all would be ok going forward.

If you do decide that leaving everything inheriting is right for most projects then you may also want to have certain projects that are more ‘secret’ and for these you will need to continue to manage the permissions and user on an individual level. One thought I had was to set a property on these ‘special’ sites via PowerShell and then you could use this property to filter out in a modified version of the above PowerShell command and ensure you didn’t reset the role inheritance accidentally. I should also point out that if you use the Synchronize option on the Project Sites page then this would re-break the inheritance – so should be avoided.

As a guide we have seen the issue with a customer with around 3000 users where they are nearly all added to each of their 1500 sites. And as promised, here are the error messages and ULS entries. Different users may see different symptoms – but the user who initiates the issue, perhaps by inactivating a couple of resources, will see the ‘Save’ button on the page apparently stick on the ‘clicked’ position and eventually get a “An unexpected error has occurred.” message. The correlation ID will be found in the ULS logs and will have several rows all relating to a SQL deadlock and the Critical level one will look like:

08/10/2011 12:17:02.85 w3wp.exe (0x2178) 0x314C SharePoint Foundation Database 5586 Critical Unknown SQL Exception 1205 occurred. Additional error information from SQL Server is included below. Transaction (Process ID 80) was deadlocked on lock resources with another process and has been chosen as the deadlock victim. Rerun the transaction. 886d9cdd-5c0c-4f3a-8f89-f4e8c92acde3

Another High level one that gives more information on the query causing the issue will be something like:

08/10/2011 12:17:06.97 w3wp.exe (0x2178) 0x314C SharePoint Foundation Database tzkv High SqlCommand: 'SET NOCOUNT ON; DECLARE @DN nvarchar(256),@LN nvarchar(128),@@DocUIVersion int,@@S uniqueidentifier,@@Level tinyint; DECLARE @ItemId int; DECLARE @@iRet int; DECLARE @ExtraItemSize int; SET @@Level = 1; SET @@S=@wssp0; EXEC @@iRet = proc_SecRemoveUserFromSite @@S, @wssp1, @wssp2 SELECT @ItemId = @wssp3 IF @@iRet <> 0 BEGIN GOTO DONE; END ;BEGIN TRAN IF NOT EXISTS( SELECT tp_ID FROM UserData WHERE tp_ListId = '06C8C9BB-B10B-4042-8859-9F9985E73E76' AND tp_ID = @ItemId AND tp_Level = 1 AND tp_RowOrdinal =0) BEGIN SELECT @ExtraItemSize = 0 EXEC @@iRet = proc_AddListItem @SiteId….

I have shortened it considerably – but the key piece is the proc_SecRemoveUserFromSite. Finally the ‘Unexpected’ one:

08/10/2011 12:17:06.97 w3wp.exe (0x2178) 0x314C SharePoint Foundation Runtime tkau Unexpected System.Runtime.InteropServices.COMException: Exception from HRESULT: 0x80131904 at Microsoft.SharePoint.Library.SPRequestInternalClass.UpdateMembers(String bstrUrl, UInt32 dwObjectType, String bstrObjId, Guid& pguidScopeId, Int32 lGroupID, Int32 lGroupOwnerId, Object& pvarArrayAdd, Object& pvarArrayAddIds, Object& pvarArrayLoginsRemove, Object& pvarArrayIdsRemove, Boolean bRemoveFromCurrentScopeOnly, Boolean bSendEmail) at Microsoft.SharePoint.Library.SPRequest.UpdateMembers(String bstrUrl, UInt32 dwObjectType, String bstrObjId, Guid& pguidScopeId, Int32 lGroupID, Int32 lGroupOwnerId, Object& pvarArrayAdd, Object& pvarArrayAddIds, Object& pvarArrayLoginsRemove, Object& pvarArrayIdsRemove, Boolean bRemoveFromCurrentScopeOnly, Boolean bSendEmail) 886d9cdd-5c0c-4f3a-8f89-f4e8c92acde3

Once the sever is in the condition – which could last 15-30 minutes, then other users will get timeouts on their pages and the ULS logs may show the following:

08/10/2011 12:20:22.30 w3wp.exe (0x1228) 0x1454 SharePoint Foundation Monitoring b4ly High Leaving Monitored Scope (ExecuteStoredProcedureDataReader -- MSP_AUTH_GETUSERBYNAME). Execution Time=120002.728838442 2be0491a-a64b-4237-8cfc-40342a374d49

08/10/2011 12:20:22.30 w3wp.exe (0x1228) 0x1454 Project Server General 8ym5 Monitorable PWA:http://<server>/PWA, ServiceApp:Project Web App Service Application, User:, PSI: SqlException occurred in DAL: <Error><Class>0</Class><LineNumber>0</LineNumber><Number>-2</Number><Procedure></Procedure> <Message> System.Data.SqlClient.SqlError: Timeout expired. The timeout period elapsed prior to completion of the operation or the server is not responding. </Message> <CallStack> at System.Data.SqlClient.SqlConnection.OnError(SqlException exception, Boolean breakConnection) at …

I should also point out that use of the Project Site Provisioning Settings page option to not automatically synchronize users may avoid you getting in to this situation – but you still need some process to control access – and if most sites are unrestricted then the inheritance option from PWA may be worth a try. Just as a reminder – the option on the Project Site Provisioning Settings page looks like this:

and un-checking will stop the automatic addition of Project Server users to sites (but will not remove ones who are already there).

Hopefully the workarounds given will assist in avoiding these types of issues if you really need to have very large numbers of users accessing each of a large number of project sites. As promised – once we have a TechNet article out in the wild I will link to it.