Ondrej Žilinec – IT Blog

Today I was solving issue with not replicating DFSR server in Full Mesh topology. Here are couple helpful commands to work with DFSR:

To look if any replication is in progress:

dfsrdiag replicationstate

To look if there is anything in backlog:

dfsrdiag backlog /rgname:Rep_Group /rfname:Rep_Folder /SendingMember:SERVER01 /Receivingmember:SERVER02

To disable debug logging:

wmic /namespace:\\root\microsoftdfs path dfsrmachineconfig set enabledebuglog=false

To set maximum debug files:

wmic /namespace:\\root\microsoftdfs path dfsrmachineconfig set maxdebuglogfiles=2000

To set maximum lines in one log file:

wmic /namespace:\\root\microsoftdfs path dfsrmachineconfig set maxdebuglogmessages=400000

To look for replication folder GUID:

wmic /namespace:\\root\microsoftdfs path dfsrreplicatedfolderconfig get replicatedfolderguid,replicatedfoldername

To clean up conflict directory:

wmic /namespace:\\root\microsoftdfs path dfsrreplicatedfolderinfo where “replicatedfolderguid=’GUID'” call cleanupconflictdirectory

That’s all folks for today,

In my class today we needed to do right-click in Windows using keyboard. I just found out that Shift + F10 simulates right-click 🙂

That’s all for today,

There is lots of great tools in Linux which are needed in Windows environment. One of the great tool from linux is “tail”. You can use it following way:

tail -f /var/log/mail.log

You will get end of the file and you see all content added to file on screen. You can view log files without need to reopen it. In Windows I use utility Trace32.exe. I was looking for some more native way to do it in Windows. There is a cmd-let Get-Content in Powershell which you can use following way:

Get-Content C:\Windows\WindowsUpdate.log -Wait -Tail 10

This tails only 10 lines from the end of the file and “waits” for new added lines. Switch “Tail” is accessible only in PowerShell 3 and higher.

Powershell is getting there,

Over the weekend I was setting up Exchange 2010 database copies using DAG. Those databases were huge (total of 3,5 TB). Line between primary and backup datacenter was 100Mbps. This migration ran over 24 hours for couple databases. When I checked the status I saw job was running 23 hours and 20 minutes:

When I have checked status in about hour and half I saw interesting time:

This happened on all databases which ran over 24 hours 🙂 So it looks like Microsoft engineers haven’t think about job running more than 24 hours 🙂

That’s all from weekend,

Where there is a problem with AD replication, there is most of the time problem with DNS. Most of the time there are bad DNS records or missing DNS records. There is cool utility to check DNSLint.exe from Microsoft. It is designed to do all manual check I do when trying to solve AD replication problems.

You can download it from here and also read more about it.

It’s bad it’s not included into operating system by default.

Couple days ago I migrated fileserver. We moved redirected My Documents for users on another server. We also hided these shares behind DFS namespace to make it more manageable in future. But we found out problem with this setup –  offline files. Windows XP kept offline copies of old and new redirected files. We had to clean this offline data database. I setup start up script:

@ECHO OFF
IF EXIST C:\OfflineFilesClean.txt (
  REM Do one thing
) ELSE (
  reg.exe add “HKLM\SOFTWARE\Microsoft\Windows\CurrentVersion\NetCache” /v FormatDatabase /t REG_DWORD /d 1 /f
  echo Vymazane > C:\OfflineFilesClean.txt
)

This script was dedicated just for Windows XP, so I had to apply GPO using WMI filter:

SELECT * FROM Win32_OperatingSystem WHERE Version LIKE “5.1%” and ProductType = “1”

This script would work also on Windows 7 and Windows 8, but those OS didn’t have problem.

That’s all for today,

We had some problems with backup system backing up Windows Server 2012 deduplication enabled volumes. We decided to disable Data Deduplication on volume. After volume was converted back into “non data deduplication mode”, we found out that all Shadow Copies meta data was lost. It means all “Previous Versions” were lost. I really don’t see any reason to act like that, but there is probably some. It would be great if Windows warns you before you do the change. But it doesn’t 🙂

Have a nice day,

When I do some change I note when and what I have done. I use notepad. I used to type date and time to every step I made, for example during migration. My colleague Robert Švec told me about key F5 which puts actual date and time into notepad document 🙂

That’s just a quick quickie today.

At one Microsoft Exchange 2010 server I was receiving too many events number 106:

This event basically describes problem with some performance counter. I had problems with about 10 performance counters and I didn’t want to deal with them separatelly. Exchange 2010 has definition its performance counters stored in directory “C:\Program Files\Microsoft\Exchange Server\V14\Setup\Perf\”. There are XML files which define performance counters. Before you can manipulate with performance counters you need to run Exchange Management Shell and run in it following command:

add-pssnapin Microsoft.Exchange.Management.PowerShell.Setup

This Snap-in allows you to use two Cmd-lets:  Remove-PerfCounters and New-PerfCounters. So let’s do two things which help us to get rid of events 106:

Deregister all performance counters for Exchange:

[PS] C:\Program Files\Microsoft\Exchange Server\V14\Setup\Perf>dir *.xml | foreach { Remove-PerfCounters -DefinitionFileName $_.Name }

Register all performance counters for Exchange:

[PS] C:\Program Files\Microsoft\Exchange Server\V14\Setup\Perf>dir *.xml | foreach { New-PerfCounters -DefinitionFileName $_.Name }

Since I reregistrered all performance counter definitions into Exchange 2010 server I have no problem with event 106.

This was quickie for today 🙂

I always don’t remember commands to list all FSMO roles in domain so I decided to take a quick note into my diary 🙂 :

• Connect to domain controller
• run ntdsutil
• write roles
• write connections
• write connect to server SERVER_NAME
• write q
• write select operation target
• write list roles for connected server

More sexy command is

netdom query /domain:DOMAIN_NAME fsmo

and viola. I know it’s dummy post, but I had to wrote it down 🙂