granth's blog : TFS at Microsoft

VSTS Pioneer TFS2010: SQL Backups

grantholliday — Fri, 13 Nov 2009 22:29:00 GMT

(See this summary on the VSTS Pioneer dogfood server and all the other posts.)

Having an adequate backup strategy for your Visual Studio 2010 Team Foundation Server is a very important part of any deployment. I’ll share with you how we are running backups on our Pioneer Dogfood TFS2010 server, but you should refer to the MSDN guidance on How to: Back Up Team Foundation Server for definitive guidance.

To prevent accidental corruption of data, you must synchronize all backups of databases that Team Foundation Server uses, which include databases for SQL Server and SharePoint Products, to the same time stamp.

In the fine print a lot of people (myself included) don’t realize that you must restore all databases to the same point in time otherwise you may not have a functioning system. The stumbling block that we ran into on a recent upgrade dry-run was that the ‘Tfs_Configuration’ database was restored to a point in time that was 3 hours earlier than one of our collection databases. This caused identity information in the collection database to be out of sync with the configuration database and caused the cloned system to be unusable.

The reason this happened is because it takes us ~3 hours to backup our collection database, so the timeline went something like this:

12:00 Backup Tfs_Configuration (1 minute)
12:01 Backup Tfs_VSTSDF collection (3 hours)

The additional step we were missing in our dry-run steps was to restore the transaction logs for all the databases to a common point in time (a time after the collection database had finished backing up). See Restoring a database to a point within a backup and Performing a Complete Database Restore (Full Recovery model).

Our backup process

SQL backups are performed using stored procedures from Ola Hallengren. See SQL Server 2005 and 2008 - Backup, Integrity Check and Index Optimization

Essentially a stored procedure is called with a number of parameters from a SQL Agent Job which performs the backups.

The reason these scripts are used instead of standard SQL Maintenance Plans, is because they give us a more flexible and robust backup solution. For instance, the backup job log is written as the job executes & we have the flexibility to exclude certain databases while automatically including all new databases created on the server. See Comparison between SQL Maintenance Plans and Ola's Backup scripts

Job Name	Location	Schedule	Retention	Databases
DatabaseBackup - Weekly FULL	E:\MSSQL\Backup\Full	Occurs every week on Wednesday at 12:00:00 AM	Delete files older than 144 Hours (6 days)	SYSTEM_DATABASES, USER_DATABASES
DatabaseBackup - Nightly DIFF	E:\MSSQL\Backup\Diff	Occurs every week on Monday, Wednesday, Thursday, Friday, Saturday, Sunday at 10:00:00 PM	Delete files older than 23 hours	SYSTEM_DATABASES, USER_DATABASES * Automatically does a FULL backup if a DIFF cannot be performed.
DatabaseBackup - 15min LOG	F:\MSSQL\Backup\Log	Occurs every day every 15 minute(s) between 12:00:00 AM and 11:59:59 PM.	Delete files older than 24 hours	SYSTEM_DATABASES, USER_DATABASES

Then we have our IT group’s tape backup process come along every night and backup the SQL backup files to tape for on-site and off-site retention.

Backup Job Alerts

Email alerts are sent to a monitoring email list when each SQL backup job completes (regardless of success or failure). This is done using the SQL Mail functionality.

Compression

Additionally, if you are running SQL Server 2008 Enterprise Edition – you have SQL Backup Compression available to you. This can greatly reduce the size of your backups and the amount of I/O required, therefore decreasing the time it takes to perform backups & restores. The tradeoff is of course, additional CPU usage. See the following SQLCAT Tuning the Performance of Backup Compression in SQL Server 2008 article for more information on this.

TFS2010: Public Workspaces

grantholliday — Sun, 08 Nov 2009 23:35:37 GMT

Public Workspaces is a feature that we built last year mainly for our internal build lab customers. Before TFS2010, if one user tries to use a workspace that belongs to another user on the same machine, they will receive an error message:

No workspace matching *;SecondUser on computer COMPUTER found in Team Foundation Server http://tfsserver:8080.

The Public Workspaces feature removes this limitation of one user per workspace and allows multiple users to use the same workspace mappings on a single machine. Unlocked workspaces are sometimes also referred to as "Shared workspaces", as they can be shared among multiple people.

Additionally, we enabled the ability to change the owner of a workspace, if you don’t want to share a workspace at all, but just reassign it to a different user.

Scenarios

Shared Machines

"I need to setup a shared ProductX enlistment machine; we’d like to use it to run tests, integrations and such."

"Our team has a rotating branch-owner role, so we have no choice but to use a service account."

"We find it very useful to have a shared machine that anyone on our team can use to investigate issues, checkin fixes, etc."

"Is there a way to make a TF enlistment user-neutral, so that anyone who logs into the build machine can open the enlistment, and use TF commands – and those commands will be run as the username of the loggedin user?"

Save disk space

"We need to use it to save disk space on shared machines."

Split personalities

"I have two admin accounts on the box. I use one of them to enlist in the branch and work with TFS. The problem is when I use another one (labaccount vs. personal account) I can login, but I cannot run any tf command because enlistment belongs to the first user only."

Hotdesking

"I'm taking over another dev's machine for a week while he works on another project and he has all the development tools already setup"

Workarounds for TFS2008

Before this feature existed, there were are few different workarounds that people were using:

Workaround	Description
Shelvesets	The various developers pass shelvesets around
Run As	The owner starts a cmd shell / VS window with RunAs
Shared User Account	Create the workspace using a shared system account, then have all developers use that account on that machine. This has the downside that all changes are checked in as the shared account, rather than the actual user. You can overcome this by using the "/author" flag with the actual username, but only if the shared account has the CheckinOther permission.
SUBST path substitution	Use the windows SUBST.EXE "Substitution" utility to map a different drive letter to the same physical path. This tricks TFS into thinking that the workspace doesn't exist.

Usage

The owner of the workspace should run the "tf workspace" command and select a workspace permission profile from the drop-down box. The default is "private."

A Private workspaces works the same as a workspace in TFS2008.
A Public (limited) workspace can be used by any valid user. Only the owner can check in or administer the workspace
A Public workspace can be used by any valid user. Anybody can check in or administer the workspace as long as they have permissions to the underlying files. The checkins will still be attributed to the person checking them in.

Workspace permissions profiles

	Private	Public-limited	Public
Read	Everyone	Everyone	Everyone
Use (most operations -- such as pend changes, resolve, etc.)	Owner only	Everyone	Everyone
Check in	Owner only	Owner only	Everyone
Administer (change owner, mappings, etc)	Owner only	Owner only	Everyone

Once a workspace has been made public or public-limited, another user can log onto the machine with the workspace and use the workspace. The other user will need to refresh their local workspace cache manually by using the "tf workspaces /collection:http://server:8080/tfs/collection" command before they will be able to use the workspace (otherwise they will not know the workspace is there -- because they did not create the workspace, and the local workspace cache is per-user).

VSTS Pioneer TFS2010 Dogfood Server: Statistics October 2009

grantholliday — Fri, 06 Nov 2009 09:21:36 GMT

(See this summary on the VSTS Pioneer dogfood server and all the other posts.)

As with the last few months, October was another busy month for dogfooding TFS. Since the Pioneer server has been such a success for Work Item Tracking and Source Control, we’ve really ramped up our efforts to dogfood every single feature as much as we can. The benefits of having everything online really lets us make sure our end-to-end integration is working as well as feel the pain before we share it with our customers. The most notable changes this month are:

Each of the different business units within VSTS have setup their own SharePoint MOSS Dashboards with bug trends, etc.
We’re now using Team Build 2010 with Gated Checkin on our part of the tree (we still have to follow the divisional processes of custom nightly build scripts at this stage)
We enabled Test Case Management features and are now dogfooding the Test Case Management (TCM) features for our major test passes
We setup and enabled the Lab Management functionality on our server and we’re starting to dogfood that across the organization as well

Here’s the latest stats:

Recent Users: 915

Version Control

Version Control Users: 622
Files: 11,053,314
Compressed File Sizes: 284,480
Uncompressed File Sizes: 870,228
Checkins: 29,275
Shelvesets: 7,249
Merge History: 10,897,164
Pending Changes: 13,344
Workspaces: 1,607
Local Copies: 382,464,960

Work Item Tracking

Users with Assigned Work Items: 2,607
Total Work Items: 284,813
Areas and Iterations: 2,910
Work Item Versions: 2,892,664
Work Item Attachments: 115,501
Work Item Queries: 5,728

Commands (last 2 weeks in October)

Work Item opens: 154,788
Work Item queries: 120,537
Work Item updates: 40,242
Checkins: 2,630
Uploads: 10,440
Shelves: 4,553
Downloads: 251,032
Gets: 41,615

DevDiv TFS Server

On top of enabling all these new scenarios on the Pioneer server, we also managed to get the main DevDiv TFS server upgraded from SQL 2005 to SQL 2008. This was the first step to getting the server to TFS2010. I can’t say it was all smooth sailing, but it looks like we’re through the worst of it now (man, what a fortnight…). The good news is that all of the extra performance optimizations we had to make to cope with our dataset sizes also made it into TFS2010 product.

Although no install guide will tell you, I highly recommend you put on SQL 2008 SP1 CU4 if you’re running a large, busy TFS2008 or TFS2010 server – in the processes of testing the upgrade and after the actual upgrade to SQL 2008, we hit these bugs which the SQL team quickly reacted and fixed for us:

KB967178 - When you restore a SQL Server 2005 backup file in SQL Server 2008, the operation takes much longer than when you restore the same backup file in SQL Server 2005 (fixed in SQL08 SP1 CU1)
KB970823 - CPU cost increases suspiciously when the number of threads working on the same index increases (fixed in SQL08 SP1 CU3)
KB970133 - Backup Log hangs when using compressed backups and the destination drive runs out of disk space. SQL Restart is required (fixed in SQL2008 SP1 CU4)

Now, onto the next Pioneer upgrade.

TFS2010: Diagnosing Email and SOAP subscription failures

grantholliday — Thu, 29 Oct 2009 01:33:04 GMT

Sometimes emails alerts such as "Work Item Changed" will fail to send. Sometimes SOAP (web service) subscriptions will not be called correctly.

To diagnose these, you can look in the tbl_JobHistory table in the Tfs_Configuration database.

If you run this SQL command, it will retrieve the last 10 failed notification jobs:

SELECT TOP 10 DATEADD(hour, -7, StartTime) as StartTimePDT, Result, ResultMessage

FROM [Tfs_Configuration].[dbo].[tbl_JobHistory]

WHERE JobId = 'A4804DCF-4BB6-4109-B61C-E59C2E8A9FF7' AND Result <> 0

ORDER BY StartTime DESC

The JobId GUID is the well-known GUID for the “Team Foundation Server Event Processing” job with the namespace: “Microsoft.TeamFoundation.JobService.Extensions.Core.NotificationJobExtension”

The DATEADD() function converts the date from UTC to local Pacific time.

ResultMessage:

There were errors or warnings during notification delivery. 0/4 emails delivered. 0/6 soap notifications delivered. 1 errors. 0 warnings. Set /Service/Integration/Settings/NotificationJobLogLevel in the TF registry to 2 to see errors and warnings.

To get detailed information about why the job failed, you will need to set a flag in the TFS registry. This is different from the windows registry and probably warrants a blog post on it’s own.

Changing NotificationJobLogLevel setting in the TFS registry using PowerShell

If you have VS2010 & PowerShell installed on your local machine and you have the appropriate permissions on the server, then you can run these steps from there. Otherwise, you will need to logon to a machine that does, or an AT. If you logon to the AT, you will need to run PowerShell as an Administrator.

To change the registry setting, replace the server url & collection name and follow these steps:

Open PowerShell
Run the following commands:

# Load client OM assembly.
[Reflection.Assembly]::Load("Microsoft.TeamFoundation.Client, Version=10.0.0.0, Culture=neutral, PublicKeyToken=b03f5f7f11d50a3a");

$collectionBaseUrl = "http://YOURTFSSERVER:8080/tfs/YOURCOLLECTION/";

$tfs = [Microsoft.TeamFoundation.Client.TeamFoundationServerFactory]::GetServer($collectionBaseUrl);
$collectionHive = $tfs.GetService([Microsoft.TeamFoundation.Framework.Client.ITeamFoundationRegistry]);

# Set the setting in the collection hive.
$collectionHive.SetValue("/Service/Integration/Settings/NotificationJobLogLevel", "2");

It’s a good idea to set this back to “0” once you’re done with it, to avoid filling up the job history logs unnecessarily.

Once you have set this setting, the next time the job runs and fails, it will log more detailed information including the exception and stack trace. For example:

There were errors or warnings during notification delivery. 0/0 emails delivered. 0/3 soap notifications delivered. 3 errors. 0 warnings. ------------------------------- Notification not delivered. Notification: WorkItemChangedEvent (DeliveryType: Soap; Address: http://eventsserver/EventHandler.svc) Exception: Microsoft.TeamFoundation.TeamFoundationServiceUnavailableException: Team Foundation services are not available from server http://eventsserver/EventHandler.svc. Technical information (for administrator): HTTP code 415: Cannot process the message because the content type 'application/soap+xml; charset=utf-8' was not the expected type 'text/xml; charset=utf-8'. ---> System.Net.WebException: The remote server returned an error: (415) Cannot process the message because the content type 'application/soap+xml; charset=utf-8' was not the expected type 'text/xml; charset=utf-8'.. at System.Net.HttpWebRequest.GetResponse() at Microsoft.TeamFoundation.Client.TeamFoundationClientProxyBase.AsyncWebRequest.ExecRequest(Object obj) --- End of inner exception stack trace --- at Microsoft.TeamFoundation.Client.TeamFoundationClientProxyBase.ProcessHttpResponse(HttpWebResponse response, Stream responseStream, WebException webException, XmlReader& xmlResponseReader) at Microsoft.TeamFoundation.Client.TeamFoundationClientProxyBase.ExecWebServiceRequest(HttpWebRequest request, XmlWriter requestXml, String methodName, HttpWebResponse& response) at Microsoft.TeamFoundation.JobService.Extensions.Core.TeamFoundationNotificationClient.Notify(String eventXml, String tfsIdentityXml, Subscription subscription) at Microsoft.TeamFoundation.JobService.Extensions.Core.NotificationJobExtension.SendSoapNotification(TeamFoundationRequestContext requestContext, TeamFoundationNotification notification, TeamFoundationIdentityService identityService)

This particular error message indicates that the client was expecting a SOAP 1.1 message, and the server sent it a SOAP 1.2 message.

If your email alerts are failing to send, then you’ll get an SMTP error back that shows you where to start looking.

VSTS Pioneer TFS2010 Dogfood Server: Upgrading

grantholliday — Mon, 19 Oct 2009 08:58:00 GMT

(See this summary on the VSTS Pioneer dogfood server and all the other posts.)

One of the major goals of the Pioneer Team Foundation Server is to upgrade it to early builds so that we can get some “bake time” with them and feel confident with what we’re releasing to customers. Since we spun up the server, we’ve done two successful upgrades and we’re already planning the next one:

July 1st – Initial installation of Beta 1 build
August 29 – Upgrade to 8/22 build from the VSTS branch
September 17 – Upgrade to 9/17 build from the Beta 2 branch (yes, the build dropped at 5PM and we used it to upgrade at 6PM)
November – Upgrade to a post-Beta2 build

The process of testing the upgrades on a copy of our database is just as important as running the upgrade itself. With our interesting dogfood topology of a dedicated SQL server, NLB/multiple application tiers, a dedicated MOSS server, virtual machines and use of virtual DNS names, we find some interesting bugs that we’re able to fix before things are locked down for shipping.

There’s no whitepaper on “How to Dogfood” at Microsoft, so our process for deploying perpetual upgrades is something that just kind of happened. Here’s basically what we do:

Depending on where we are in the cycle and the amount of churn in the code base, we might add another iteration or skip one. What’s most interesting about this process, is that it’s driven by the end date. Based on the product schedule, we know when our windows of opportunity are for upgrading and we work back from then.

Tips for upgrading

Here are some tips to consider when planning your own upgrades:

Hardware – Get your production & pre-production servers in place early. Getting new hardware or re-configuring existing hardware usually has a long lead time. Make sure that you’ve also got enough storage space.
Start fresh – If at all possible, give yourself the best chance at success by starting with a fresh box. Sure, plenty of products support in-place upgrades & clean uninstalls – but in my experience, flattening the box and starting fresh lays a good foundation for the future.
Backups – Organize a copy of your production backups that you can test the upgrade with. Not only does it give you peace of mind that your backups work, it also helps you line up the DBA’s for when it’s time to do the real upgrade.
Start with the latest OS – If you start with Windows 2008 R2 now, it’s one less OS upgrade (and downtime) you’ll have to do in the future. It also means you’ll be on 64-bit, which is a very good thing.
Virtualization – We run our 3x Application Tiers as virtual machines with no problems whatsoever. This is great for when we’re installing pre-beta copies of the .NET framework that don’t have clean uninstalls yet. In our last upgrade we just turned off the old application tiers and spun up three fresh ones. You can treat AT’s as throwaways in a TFS2010 NLB topology, since there is no state stored on them.
Dry-Runs – Run through your plan and make sure everything is going to work. Review it with others and get their feedback.
Have a test plan – Once you’ve done the upgrade, you need to be sure that everything is back online and working as expected. Checkin, Checkout, Open work item, Save work item, Web Access, Sharepoint, Reports, Cube, Alerts, Monitoring, Builds, etc.
Do pre-requisites ahead of time – For example, to install SQL2008 you need to have .NET3.5 and Windows Installer 4.5 both installed first and both of these require a reboot. If you plan to do one on the first weekend and the other on the second weekend, it means that you don’t waste time during your more important SQL upgrade waiting for reboots.
Do one thing at a time – This is an extension of the last point. Let’s say you’re moving from a single-server TFS2008 SP1 on SQL2005 and Windows 2003 to a dual-server TFS2010 on SQL2008 SP1 and Windows 2008 R2. If it were me, I’d take it in steps: upgrade SQL, make sure everything still works, move to dual-server, check again, upgrade to TFS2010. The more steps you try and bundle together, the higher your risk of failure.

Upgrade logistics

I’ve been involved with at least 5 major upgrades this year and here are some things I’ve learnt from running them:

Book a conference room
- This is the single most valuable thing you can do. Try and get anybody who is actually doing any steps in the upgrade (DBA< IT, Helpdesk, etc) to come to the room.
- Have a spare computer or project your own desktop on the screen. This is useful for synchronizing with others in the room, running ‘ping –t’ when a server reboots and doing “group debugging” sessions when problems happen.
Setup a “party line” conference call. Invite anybody who wants to “listen in” to join. Invite your manager.

It’s also useful for anybody who you might need help from to have these details. If things go south, you don’t waste time trying to setup a conference call – it’s already there.

Setup a live meeting / desktop sharing session. Invite anybody who wants to be a “fly on the wall” to join.
- Sure, you can log a bug that says “when I clicked on X on screen Y, setup blew up” and people might believe you. But if you’re sharing your desktop and other people see it happen, then you have some people to back you up.
- (I’ve been meaning to do a screen recording of our upgrades, but various technology problems have prevented it each time – I think it would be useful for reviewing and improving the process for next time)
Setup an “on call / escalation” list. Find out who you need to call if you have problems with one of your dependent systems. The last thing you want to be doing is sitting on hold to a L1 helpdesk if your server doesn’t reboot cleanly or something.
Send at least 3 notifications to users of the downtime
- 1st Notice – Give as much notice as possible so people can plan milestones around the date. Even if it’s not set in stone yet, people would rather know about it and have it cancelled than to find out at the last minute.
- 2nd Notice – 7 days or on the Monday before a weekend upgrade is good for this.
- 3rd Notice – On the day of the upgrade (or 24 hours before if it’s a weekend), send another one.
- Guidelines: Make sure you describe what the impact to users is, any actions they need to take and where they can go to get future status updates and support.
Provide regular status to stakeholders. Start an email thread and trickle through progress as it happens. By keeping them informed, you buy their trust and their support when things don’t go as planned.

Planning your own upgrade

Brian Keller has a great post on how to Get ready to “go live” with Team Foundation Server 2010 beta 2! He includes two useful documents:

Overview presentation (.PPTX)

Detailed checklist (.DOCX)

I’m also going to share with you a copy of the generic deployment plan template that I’ve been using for the Pioneer upgrades.

Deployment Plan Template(.XLSX)

Planning, testing and upgrading to these new releases has been a fun experience and it’s a bonus when everything goes smoothly. It’s great to see our experiences influence the product and make it the best release yet.

VSTS Pioneer TFS2010 Dogfood Server: Statistics September 2009

grantholliday — Mon, 19 Oct 2009 07:13:18 GMT

(See this summary on the VSTS Pioneer dogfood server and all the other posts.)

September was a busy month for dogfooding TFS – all of the VSTS team were completely moved across and on 9/17 we upgraded to an almost final TFS2010 Beta 2 build. Here are the numbers from the end of September:

Recent Users: 887

Version Control

Files & Folders: 10,722,118
Compressed File Sizes (MB): 263,350
Uncompressed File Sizes (MB): 773,731
Checkins: 23,369
Pending Changes: 55,697
Local Copies: 303,331,190

Work Item Tracking

It looks like I don’t have the work item tracking stats for September – I’ll have them next month.

Commands (last 2 weeks in September)

Work Item opens: 113,128
Work Item queries: 101,203
Work Item updates: 23,653
Checkins: 1,855
Uploads: 6.622
Shelves: 2,402
Downloads: 216,452
Gets: 30,201

Table sizes

One of the interesting metrics that I’m also tracking is how our tables and indexes are growing on the disks. Here’s the top 10 for this month:

Table	Rows	Reserved (MB)	Data (MB)	Index (MB)	Unused (MB)
tbl_Content	5,915,547	277,260	275,129	4	2,127
tbl_LocalVersion	303,331,190	81,562	44,789	36,574	199
Attachments	63,783	11,928	11,908	13	7
tbl_Version	12,129,564	3,418	1,031	2,369	17
WorkItemLongTexts	2,728,618	3,314	2,533	777	5
tbl_VersionedItem	15,754,176	2,983	1,598	1,382	3
WorkItemsWere	1,239,831	2,482	1,617	863	2
tbl_PropertyValue	33,293,914	1,320	523	793	4
tbl_File	5,717,981	722	442	266	14

TFS2010: Installing pre-requisites from the command line

grantholliday — Mon, 19 Oct 2009 06:31:35 GMT

On the Team Foundation Server team, there is a new build of the server every night that comes out of our build lab. These vary in quality throughout the development cycle and the QA teams look at them and rate them as “Self Test”, “Self Test (with workarounds)”, or “Self Toast”. The QA & dev guys have lots of fancy automation for deploying nightly builds and testing them out, but I like to stay away from that stuff and use the builds as if I were a real customer. Once a build is declared “Self Test”, I’ll grab a copy and spin it up in a VM to see what’s changed.

The two preparation things I’ve done that have really sped up my installation turnaround times are:

Create a sysprep’d Windows Server 2008 R2 VHD
Create a SQL2008 SP1 slipstreamed installation point

Once I had those in place, I then just used these two long command lines below to install SQL & IIS. Of course, none of this is necessary for a “TFS Basic” install, since the installer will drop down IIS & SQL Express for you. See Brian’s blog on TFS 2010 for SourceSafe Users for more details on that.

Creating a sysprep’d Win2008 R2 VHD

Creating a sysprep’d VM is easy. Create a new VM and install a fresh copy of Windows 2008 R2. If you have a domain, join it to the domain and add yourself to the local administrators group (this will be retained after you sysprep & rejoin to the domain again). I don’t usually add the IIS role at this point, since I use these VHD’s for other things. Once you’re happy with it, run the following command:

C:\Windows\System32\Sysprep\Sysprep.exe /oobe /generalize /shutdown

Then take a copy of the VHD and name it something like ‘Win2008R2Sysprepped.vhd’. Whenever you need a fresh machine, just copy this VHD to a new file and away you go.

Creating a SQL2008 SP1 slipstreamed installation point

Follow the instructions for creating a slipstreamed drop on the SQL Server Setup team’s blog. Two things to note when doing this:

Paths matter – it’s best to pick a folder name and stick with it whenever you use it. e.g. C:\SQL2008SP1\
You DO have to download the ia64 & x86 SP1 packages & slipstream them as well, otherwise your setup.exe will fail

Once you’ve created the installation point, save it away somewhere – it’s really useful and it saves so much time by not having to do an RTM install followed by a SP1 install.

Installing TFS pre-requisites from the command line

Boot into your fresh 2008 R2 VM
Give it a new name, join it to the domain, apply any updates, reboot, etc
Install SQL2008 SP1 from your installation point:

Open a command prompt as administrator
Type: pushd \\fileserver\SQL2008SP1
Type the following command:

Setup.exe /q /ACTION=Install /FEATURES=SQL,FullText,AS,RS,Tools /INSTANCENAME=MSSQLSERVER /SQLSVCACCOUNT="NT Authority\Network Service" /RSSVCACCOUNT="NT Authority\Network Service" /ASSVCACCOUNT="NT Authority\Network Service" /AGTSVCACCOUNT="NT Authority\Network Service" /SQLSVCStartuptype=2 /RSSVCStartupType=2 /ASSVCSTARTUPTYPE=2 /AGTSVCSTARTUPTYPE=2 /BROWSERSVCStartupType=2 /SQLSYSADMINACCOUNTS="Builtin\Administrators" /ASSYSADMINACCOUNTS="Builtin\Administrators" /SQMREPORTING=1 /ERRORREPORTING=1 /INDICATEPROGRESS

Install required IIS components:

Open a command prompt as administrator
Type the following command:

ServerManagerCmd -install WAS WAS-Process-Model WAS-NET-Environment WAS-Config-APIs Web-Server Web-WebServer Web-Common-Http Web-Static-Content Web-Default-Doc Web-Dir-Browsing Web-Http-Errors Web-Http-Redirect Web-App-Dev Web-Asp-Net Web-Net-Ext Web-ISAPI-Ext Web-ISAPI-Filter Web-Health Web-Http-Logging Web-Request-Monitor Web-Security Web-Windows-Auth Web-Filtering Web-Performance Web-Stat-Compression Web-Mgmt-Tools Web-Mgmt-Console Web-Mgmt-Compat Web-Metabase Web-WMI Web-Lgcy-Scripting Web-Lgcy-Mgmt-Console

At this point, you’re now ready to run setup.exe from the TFS media. You have a fully configured SQL server and the requires IIS pre-requisites installed.

Other useful SQL 2008 setup.exe command lines

For a TFS “Application Tier Only” install, IIS is one pre-requisite and the “SQL Client Tools Connectivity” is the other one. This is required so that the TFS Job Agent can talk to Analysis Services and tell it to process the warehouse. Here’s how to install just the client tools:

setup.exe /ACTION=Install /FEATURES=Conn /ERRORREPORTING=1 /INDICATEPROGRESS /SQMREPORTING /Q

If you just want to install SQL Server Management Studio on your developer machine:

setup.exe /ACTION=Install /FEATURES=ADV_SSMS /ERRORREPORTING=1 /INDICATEPROGRESS /SQMREPORTING /Q

For more information, see How to: Install SQL Server 2008 from the Command Prompt and available features.

Scripting Team Project Creation in TFS2010

grantholliday — Sat, 26 Sep 2009 00:17:46 GMT

In TFS2008 SP1 and the subsequent power tools release it was possible to create a team project from the command line. You can read Will-Peter’s detailed blog post on Creating a Team Project from the command line in TFS2008.

In TFS2010, it’s still not as good as we would like (it’s still client-based rather than server-based). But there’s one important difference – the feature has been rolled into the product and is no longer a power tool.

How to create a team project from the command line in TFS2010:

Create an XML file with the following contents and save it somewhere (e.g. C:\temp\pcw.xml). Substitute the highlighted as appropriate:

<?xml version="1.0" encoding="utf-8"?>
<Project xmlns="ProjectCreationSettingsFileSchema.xsd">
       <TFSName>http://tfsserver:8080/tfs/DefaultCollection</TFSName>
       <LogFolder>c:\temp</LogFolder>
       <ProjectName>TestAgile2</ProjectName>
       <ProjectSiteEnabled>true</ProjectSiteEnabled>
       <ProjectSiteTitle>TestAgile2</ProjectSiteTitle>
       <ProjectSiteDescription>TestAgile2 project created from command line</ProjectSiteDescription>
       <SccCreateType>New</SccCreateType>
       <SccBranchFromPath></SccBranchFromPath>
       <ProcessTemplateName>MSF for Agile Software Development v5.0</ProcessTemplateName>
</Project>

Once you have created the file, you need to prepare Visual Studio to be automated and trigger the automation from the command line.

Open Visual Studio 2010
Go to View > Team Explorer and make sure that the Team Explorer tab is active
Close Visual Studio 2010
Open a command prompt
Type: CD /D %ProgramFiles(x86)%\Microsoft Visual Studio 10.0\Common7\IDE
Type: devenv /command "File.BatchNewTeamProject C:\temp\pcw.xml"

At this point, Visual Studio will open and the Team Explorer tab will show “Connecting to server\collection…”. In the background it’s actually creating the team project. You can verify this by looking for a txt file with the name of your project in the path you specified for LogFolder. (e.g. c:\temp\TestAgile2.txt)

Once the project has been created, you can select ‘Add existing team project’ and select it from the dialog.

If you want to create a fully-automated project provisioning system there are a few other things that you’ll need to do on top of creating the team project. Some of these things are:

Add users to TFS groups (Project Administrators, Contributors, Readers)
Add permissions in sharepoint
Add permissions in reporting services

VSTS Pioneer TFS2010 Dogfood Server: Statistics August 2009

grantholliday — Thu, 27 Aug 2009 21:44:00 GMT

(This blog post is part of a series of posts on the new VSTS Pioneer TFS dogfood server.)

Everybody loves statistics, so here are the statistics for the first month on the VSTS Pioneer server.

Recent Users: 415

Work Item Tracking

Users with Assigned Work Items: 2,454
Work Items: 111,319
Areas & Iterations: 2,811
Work Item Versions: 1,056,416
Work Item Attachments: 50,833
Work Item Queries: 2,965

Version Control

Files & Folders: 8,561,565
Compressed File Sizes (MB): 325,080
Uncompressed File Sizes (MB): 856,859
Checkins: 18,867
Shelvesets: 1,982
Merge History: 1,789,333
Pending Changes: 48,155
Workspaces: 727
Local Copies: 184,231,208

These statistics are captured by using the TFS Best Practices Analyzer and a custom configuration file. The configuration tells the BPA tool to run a series of SQL queries and dump the results to an XML file. I have then created an XSLT that transforms the results into a more user-friendly HTML file. The BPA tool is then scheduled to run every night and generate a fresh set of statistics. I’m working with the TFS BPA team to get this configuration built into the next version of the BPA tool – I don’t have an ETA on when that will ship, but stay tuned.

This month I’ve also pulled the last 30 days of PerfMon counters to see how our hardware choices are holding up. If you’d like to setup something like this for yourself, have a look at my previous posts.

Disk Performance

Things are looking OK on the IO front. 95% of the time we’re pushing 45MB/sec (Read) and 33MB/sec (Write) or less to and from our storage. The response time is <40ms which is nothing to worry about for the configuration we have.

LogicalDisk(_Total)	50th Percentile	95th Percentile	99th Percentile
Avg. Disk sec/Read	0.004	0.039	0.371
Avg. Disk sec/Write	0.003	0.023	0.063
Disk Read Bytes/sec	2,142,548	44,926,021	94,273,782
Disk Write Bytes/sec	1,908,518	33,029,252	70,802,346
Avg. Disk Bytes/Transfer	42,867	133,887	436,920

Memory

The memory usage on our SQL server has remained completely constant. This is because SQL is set to use a fixed amount and Analysis Services & Reporting Services share the rest.

CPU Performance

(Avg: 32%). The majority of this usage is SQL Server, with a small portion going to Analysis Services & Reporting Services. It’s nice to see that we still have plenty of capacity left for the usage to grow.

That’s all for statistics, stay tuned for more next month.

VSTS Pioneer TFS2010 Dogfood Server – Hardware & Topology

grantholliday — Thu, 27 Aug 2009 18:23:00 GMT

(This blog post is part of a series of posts on the new VSTS Pioneer TFS dogfood server.)

While running the main TFS server for ~3,500 users in Developer Division we got a pretty good idea of the hardware required to support our enormous number of files, local versions & branches.

We knew that:

our SQL Server load patterns were very IO intensive
our SQL Server load patterns were very CPU intensive
our disk usage will grow by about 300% per year
the number of operations the server performs will increase

Servers

We wanted to purchase and setup hardware that was flexible enough and designed to scale as our needs do. For this reason we settled on two identical servers that are part of the Microsoft IT recommended systems for our needs:

HP ProLiant DL580G5 (room to add two more processors & upgradeable to 256GB RAM)
Windows 2008 x64 SP2
2x 6-core Intel Xeon X7460 2.66Ghz processors
32GB RAM
QLogic QLE2460 4Gbps Fiber Channel Host Bus Adapter (SAN connection)

According to Joe Chang who studies SQL Server performance, the “The X7460 (Dunnington) is a clear winner at 4-way for the high-call volume apps.” The 6-core CPUs also offer a pretty significant performance boost over the older quad-core CPUs:

“On TPC-C, the 7460 six core generated a 34% edge over the quad core AMD and 56% advantage over the quad core X3750”, and
“On TPC-E, the six core showed a 49% edge over the older quad core”

Virtual Machines

We are running 5 virtual machines at the moment with almost identical configurations. Since we’ll be constantly upgrading things like the .NET CLR that often don’t support build-to-build upgrades or clean uninstall, running the application tiers as VM’s makes perfect sense. It also improves our rollback story if we have a failed upgrade. (More about upgrade in a future post).

VM	Purpose	Configuration
AT01, AT02, AT03	TFS Application Tier	Windows Server 2008 R2 4 Virtual processors 4GB RAM 2 Virtual NICs 1 VHD for C$ 1 VHD for version control cache
MIRROR	TFS-TFS Connector tool runs here	Windows Server 2008 RTM 4 Virtual processors 4GB RAM 1 Virtual NIC 1 VHD for C$ 1 VHD for version control temporary workspace
MOSS	Microsoft Office SharePoint 2007 SP2 (Complete install) Web Front End, Excel Services, Search Services & Shared Services Provider	Windows Server 2008 R2 4 Virtual processors 4GB RAM 1 Virtual NIC 1 VHD for C$ Databases reside on physical SQL server

Storage

Since these servers are hosted by the DevDiv IT guys, we went with their recommendation for storage. They already have a massive storage infrastructure for build & drop servers built on Xiotech technology, so it was easy for them to add additional ISE units to meet our size & performance requirements. In the end we ended up with the following configuration:

5TB of “Balanced” storage
5TB of “Performance” storage

The SQL Customer Advisory Team’s – Storage Top 10 Best Practices is a great resource when planning your own TFS deployment. In particular we specifically considered the following:

Isolate log from data at the physical disk level (e.g. O$ drive is on separate spindles)
Consider configuration of TEMPDB database (e.g. T$ drive is on separate spindles)
Make sure to give thought to the growth strategy up front (e.g. We can add additional ISE units at any time and grow any drive without having to migrate data)
Lining up the number of data files with CPU’s has scalability advantages for allocation intensive workloads. (e.g. we have 12 datafiles for TempDB because we have 12 CPU cores)

Here’s how the virtual disks (LUNs) are configured on the server:

SERVER1 (SQL, SSRS, SSAS)

Drive	Purpose	Storage Type	Size (Gb)
E$	SQL backup dump drive	Balanced	2741
G$	LocalVersion filegroup	Performance	1673
I$	OLAP data	Balanced	25
J$	All database data	Performance	1833
K$	Version filegroup	Performance	971
O$	Transaction Logs	Balanced	1500
T$	TempDB	Performance	500

SERVER2 (Hyper-V Host)

Drive	Purpose	Type	Size (Gb)
E$	Hyper-V VM's	Balanced	743

What’s cool about the Xiotech stuff is that it’s like Lego blocks – as you need more storage or more performance, you just buy more blocks and plug them in. What’s even cooler about it, is that you can control the SAN using web services – the build lab uses the SANMan tool they built to move virtual disks between build servers and automate the provisioning of disk space.

Topology

At just over 400 users, we probably don’t need three AT machines in a NLB configuration. However, NLB is an important new feature of TFS2010 and it’s a scenario that we need to dogfood for ourselves.

Here’s some quick facts about our topology:

We have a couple of “friendly name” DNS records for the NLB cluster IP, Reporting Services & Analysis Services – this allows us to change the underlying infrastructure without users having to connect to a new addresses.
We’re running Windows 2008 R2 RTM on our VM’s & soon on our physical servers
A single VM was created, then sysprep’d and copied. This means that we can spin up a new VM in a very short time period. (Copy VHD, Add to Hyper-V, Start, Join Domain, Apply Updates, Done.)
SERVER2 is running an SMTP server that relays mail on behalf of each of the other servers. At Microsoft, the corporate mail servers will only accept mail from authenticated users that have mailboxes. By running the SMTP server as an authenticated user, it allows us to run the application tiers as Network Service. This means when we have to change a password to a service account, it doesn’t cause a TFS service interruption – just a brief SMTP interruption.
Since all the SQL services are running on the same box, they don’t need to be run as a domain accounts either – no service interruption & less places that need a password changed when it expires.

That’s it for hardware and topology, next post is on performance and Dogfood statistics.

TFS Performance Report Pack now works with SSRS2005

grantholliday — Wed, 24 Jun 2009 06:40:14 GMT

Back in February, I released some of the reports that were created for our internal TFS servers on my blog as the TFS Performance Report Pack. I got countless emails and comments on the fact that some of them didn’t work with SQL Reporting Services 2005.

Fortunately, Jim Saunders from the Developer Support team at Microsoft (who have a great blog!) has re-created the three reports in the pack that required SQL2008 Reporting Services. Just download a different version of the reports from here and replace the ones in the original zip file.

From the support team blog:

...He lists SQL Server 2008 Reporting Services as a requirement, with a note that they should work in SSRS 2005. It turns out some of the reports function under both SSRS 2008 and 2005, some do not. Several posters have mentioned incompatibilities with some reports and SSRS 2005.

These reports are compatible with either SSRS version:

Server Status - Historical Performance Trends.rdl

Server Status - Recent Performance Trends.rdl

Server Status - Top Users Bypassing Proxies.rdl

These reports require SSRS 2008:

Execution Time Summary.rdl

Execution Time for User.rdl

Server Status - Source Control Request Queue.rdl

I have created versions of the three reports above that are functionally similar to Grant's reports, but also work with SSRS 2005. You can download my updated reports here. To use my reports, install Grant's TFS Performance Report Pack, and replace the three reports with the versions in the zip file I provided.

TFS Adoption at Microsoft – May 2009

grantholliday — Wed, 24 Jun 2009 01:26:00 GMT

As promised last week in my Dogfooding TFS show on the RadioTFS podcast, here is the latest on TFS within Microsoft that myself and Brian have posted about in the past.

May had the all-time peak in the total number of unique active users with 15,613. In addition, the total number of source code files stored surpassed 60 million. The total number of projects continues to increase and will end the fiscal year over 3,500, growing by over 1,300 in the last year. MSIT RXD continues to lead the way using a TFS2010 pre-beta1 build with an average of 100 active users per week.

Summary of growth indicators for the month:

May 2009	Growth	Totals
Unique Active Users	+29	15,613
Total Active Users	+24	17,680
Projects	+98	3,434
Work Items	+170,436	3,934,238
Source Code Files	+1,760,516	60,253,062

We call the internal TFS at Microsoft the “TFS Service Offering”. All the servers are setup and maintained by our partners in Microsoft IT.

Active Users - Service Offering:

Projects & Unique Active Users Across Service Offering:

The number of team projects was up by 98 this month. Added to the chart this month are labels for the releases since January 2007 to give some perspective on growth and when versions were released.

Team Builds:

Combined data from across the Service Offering.

RadioTFS Podcast on Dogfooding TFS

grantholliday — Wed, 24 Jun 2009 01:22:00 GMT

Last week I sat down (virtually) with two of our Team System MVP’s - Martin Woodward and Paul Hacker. We had a good chat about TFS at Microsoft which you can listen to on show #22 of the RadioTFS podcast.

Play now: Radio TFS #22: Dogfooding TFS with Grant Holliday (35minutes, 16MB MP3)
Subscribe now: Radio TFS – A show about all things related to VSTS and TFS

In the show we talked about a range of topics. Here are some links:

Microsoft Process Template (MPT) on Codeplex
Book - The Build Master: Microsoft's Software Configuration Management Best Practices
Slot Mode vs Item Mode (aka Schema Change)
TFS Performace reports pack
Report VS2010 Issue on Connect

Large Team Foundation Server Performance Characteristics

grantholliday — Thu, 21 May 2009 20:27:42 GMT

Since joining Microsoft, I’ve become intimately familiar with running a TFS server for ~3,500 users in Developer Division and the performance characteristics of it.

One thing I’ve learnt is that Performance Counters rule. You might observe the server being “slow” and you might notice that it “takes a while” to do certain operations – but you need evidence to back up your claims before anybody will take you seriously. The evidence that everybody has access to, is reliable and people take seriously are the perf counters built into windows.

See my previous posts on Relogging Perfmon binary log files to SQL and Querying Perfmon data from SQL for more details on setting up performance counter logs.

If I think about the problems we’ve overcome in the last 12 months, the issues come down to these:

IO – If the “LogicalDisk\Avg. Disk sec/Transfer” perf counter for any of your disk drives is more than 0.030 (30ms) – then you’re hosed. This counter is a primary indicator of disk latency. Get that fixed before doing anything. (see below for more details)
Workspace Mappings – If you have unnecessary paths in your workspace mappings, then Get() will be much slower than it needs to be. E.g. DON’T map $/ to C:\Code and think that everything will be good. A root mapping isn’t truly a bad thing but if you aren’t careful it can lead to unexpected and potentially slower results.
Latency/Download requests – Proxy servers help here by offloading Download() requests from the main server. Doesn’t help Work Item tracking.
CPU – Processor performance isn’t linear. If you’re running higher than ~70% CPU for periods of time, then you need to increase your processing capacity.
SQL indexes/fragmentation – Sometimes the TFS SQL Jobs that update statistics & rebuild/reorganize indexes stop running, or don’t run for whatever reason. Check that the SQL jobs are running successfully and check for index fragmentation.

The tools you can use to diagnose performance issues are:

PerfMon. Setup a perfmon counter log for the important counters. Track them and work out what’s “normal” for your load/environment
TfsActivityLogging database. Dive into this database and look for trends, heavy users, heavy tools, etc. Understand where your load is coming from.
- Download & install my TFS Performance Report Pack and look at the Execution Time Summary report.
TfsServerManager.exe (Comes with the Team Foundation Server Power Tools, see Brian’s blog for more details)
- No shortcut gets created. Run it from "C:\Program Files\Microsoft Team Foundation Server 2008 Power Tools\TfsServerManager.exe"
- If users are reporting a problem, try and catch it while it’s currently executing. Look at the “Source Control Request Queue” report. Is their request on top?
- Run the following query a few times in SQL to see if any blocking is occurring. If the same spid hangs around for a while, run DBCC INPUTBUFFER(spid_here) to see what stored procedure it is and try and match that to a TFS command. e.g. prc_Get = Get()

SELECT a.status, a.*
FROM    sys.sysprocesses a
WHERE   spid > 50
        and spid <> @@spid
        and blocked = 0
        and EXISTS ( SELECT *
                     FROM   sys.sysprocesses b
                     WHERE b.blocked = a.spid)

Disk Latency

To determine if you are having significant issue with disk latency you should use the following performance counters:

Object: [Physical Disk] or [Logical Disk]
Counter: [Avg. Disk Sec/Transfer]
Instance: Ideally you collect this for individual disks however you may also use [_Total] to identify general issues. If [_Total] is high then further collections can be taken to isolate the specific disks affected.
Collection Interval: Ideally you should collect at least every 1 minutes. The collection should be run for a significant period of time to show it is an ongoing issue and not just a transient spike. 15 minutes is minimum suggested interval.
Issue Thresholds (seconds):
- < 0.020: Normal time and no I/O latency issues are apparent
- > 0.00 – 0.050: You may somewhat concerned. Continue to collect and analyze data. Try to correlate application performance issues to these spikes
- > 0.050 – 0.100: You are concerned and should escalate to SAN administrators with your data and analysis. Correlate spikes to application performance concerns.
- > 0.100: You are very concerned and should escalate to SAN administrators. Correlate spikes to application performance concerns.

Announcing TFS Performance Report Pack

grantholliday — Wed, 04 Feb 2009 03:00:00 GMT

Update 6/23/2009: Due to popular demand, Jim from the Developer Support team at Microsoft has re-created the three reports that required SQL2008 Reporting Services. See the support team blog for more details.

I’m on the team that runs the busiest Team Foundation Server at Microsoft. The Developer Division instance has over 3,500 active users and processes over 10,000,000 source control and work item tracking requests per day (not including the requests that our 5 proxy servers handle).

We have a responsibility to ensure that the server is performing as expected and to identify any efficiencies that can be made in the server or the tools. To do this, we have created a number of reports that we use ourselves and make available to our own users.

Download

We’re now making these reports available to you to install on your own server and monitor your TFS server’s performance. The plan is to eventually roll these into a power tool release or a subsequent release of VSTS but I’ll let you know more about that when it happens.

Download from here TFS Performance Report Pack.zip (~250Kb)
Download the SQL Server 2005 compatible reports from here SRS2005CompatibleReports.zip (~13Kb)

Requirements

SQL Server Reporting Services
A user with read-only access to the TfsActivityLogging database
A shared datasource to connect the report to (see below)
Command Logging enabled - This logging is enabled by default in TFS 2008.

In TFS 2005 you will have to enable the commandLogging flag in the Global Web.Config file (and the query for the report will be a little different).

Installing

The report pack consists of a ZIP file with a number of Report Definition (*.rdl) files. These files are designed to be deployed onto your existing TFS reporting server, e.g. http://your-tfsserver/Reports/

As a TFS administrator, extract the files to your PC
Open http://your-tfsserver/Reports/
Create a new folder in Reporting Services called “Server Status”
Create a new shared data source called “TfsActivityReportDS” and set the connection string to:
1. Data Source=localhost;Initial Catalog=TfsActivityLogging
2. Credentials: domain\user that has access to the TfsActivityLogging database
3. Use as windows credentials when connecting to the data source

Here’s an overview of what the reports look like and what questions you can answer with them.

Execution Time Summary

This report visualizes the load, in this case reflected by total execution time, on the server from two axis: users and commands.

Use this report when you want to know:

Which commands account for the largest load on the server?
Which tools / or users are putting the biggest load on the server?

Server Status - Source Control Request Queue

Source Control is undoubtedly the application that consumes the most resources on an Application Lifecycle Management (ALM) Server. Across the day, a series of requests get queued to be processed as transactions are committed to the database. This report provides a view into that queue.

Use this report when you want to know:

If a request is blocking source control operations and for how long
How healthy is the performance of version control on this hardware?

Lots of red means that you have some long running operations and you may have some problems.

Server Status - Top Users Bypassing Proxies

In the past, I’ve blogged a query to get this information – How many user’s are not using a TFS Proxy server? Internally, we have setup a scheduled subscription that emails this report twice a week.

IT departments strive to provide the best level of service to their users. Hardware requirements planning and setting up proxies are activities that ensure optimal performance for their internal teams when interacting with team Foundation Server. This report allows administrators a view into which users are not complying with internal guidelines and hence decreasing overall server performance.

Server Status - Historical Performance Trends

This report serves as a summary of the average response time for two of the Team Foundation Server subsystems: Work Item Tracking and Version Control

Use this report when you want to know:

How long are users, on average, waiting for a subsystem to process their request
Which days of the week are the most critical when it comes to performance

Server Status - Recent Performance Trends

This report provides more data granularity about the performance of the server. We start with a view into the server average response time, now looking at the entire picture instead of broken down by subsystem. We then follow with charts relating information about version control downloads and average response time distributions for the same time period.

Use this report when you want to know:

The correlation between degraded server performance and average response times by the subsystems
How does a large number of downloads affect overall server performance
Overall health indicator of the server

I hope that you find these reports useful. Please send any questions or feedback as comments to this post, or contact me via email.

granth's blog : TFS at Microsoft

VSTS Pioneer TFS2010: SQL Backups

TFS2010: Public Workspaces

VSTS Pioneer TFS2010 Dogfood Server: Statistics October 2009

TFS2010: Diagnosing Email and SOAP subscription failures

VSTS Pioneer TFS2010 Dogfood Server: Upgrading

VSTS Pioneer TFS2010 Dogfood Server: Statistics September 2009

TFS2010: Installing pre-requisites from the command line

Scripting Team Project Creation in TFS2010

VSTS Pioneer TFS2010 Dogfood Server: Statistics August 2009

VSTS Pioneer TFS2010 Dogfood Server – Hardware & Topology

TFS Performance Report Pack now works with SSRS2005

TFS Adoption at Microsoft – May 2009

RadioTFS Podcast on Dogfooding TFS

Large Team Foundation Server Performance Characteristics

Disk Latency

Further Reading

Announcing TFS Performance Report Pack

Download

Requirements

Installing

Execution Time Summary

Server Status - Source Control Request Queue

Server Status - Top Users Bypassing Proxies

Server Status - Historical Performance Trends

Server Status - Recent Performance Trends