Eoin Campbell's questions

We use Active Directory B2C to allow our client users to login to our web app.

Recently we've had users at 2 different client orgs who have been unable to login. They were able to demonstrate the following.

1. They click the login button on our website @ https://www.mycompany.com
2. They get redirected to our IdentityServer @ https://id.mycompany.com
3. They get further redirected to https://login.microsoft.com

what should happen

• The user should now go through a normal login flow
  • supplying their username ([email protected]) which was previously setup in our ADB2C
  • supplying their password which they setup during invite process
• get redirected back to our identity server
• get redirected back to our web application

what's actually happening

• The user enters their email address ([email protected])
• Microsoft immediately steps in with the following message

It looks like you're trying to access a resource that belongs to an organization that's not approved by your IT department.

What I think is happening is that because Client.com are using Microsoft Azure Active Directory/Office 365 themselves, their administrative team have set something up that tells Login.Microsoft

"We own the domain 'client.com' and if anyone ever tries to use an [email protected] mail address as the identity of a login attempt to some other tenancy, you should block them"

Is this a "feature" of AD? Clearly, this is something we need to take up with the IT department of the client company, but I'd prefer to go into that conversation armed with better information about what feature we're asking them to white list us in.

Over the weekend (Sunday March 27th) Daylight Savings Time kicked in here in Ireland and so all the servers (all win2k3 server) in our Production Environment updated their Times +1 Hour. However something which I cannot explain caused 2 of our cluster nodes to update @ 02:08 and 02:38 rather than @ 00:59.

We have 2 Domains in our production environment.

• DOMAIN1 contains PDC1 + 6 or 7 Virtual Machines. (Standalone Web, Load Bal, File servers)
• DOMAIN2 contains PDC2 + 2 Physical Servers which are clustered together for SQL Server

In DOMAIN1, everything seemed to work perfectly with all machines moving forward 1 hour at exacly 00:59:59 -> 02:00:00. This can be confirmed by viewing timestamps in logfiles on any of those servers, or viewing timestamp data written to DB tables by those servers.

In DOMAIN2 however, something odd happened.

• Jobs in the SQL Server Agent Instance on DB_NODE_1 show the time change occuring @ 02:08. i.e. there are per minute jobs which show start times of 02:08, then 03:09
• Jobs in the SQL Server Agent Instance on DB_NODE_2 show the time change occuring @ 02:38. i.e. there are per minute jobs which show start times of 02:38, then 03:39

Same goes for DB Inserts on both instances that relied on the SQL GetDate() function. Data shows the getdate() inserts jumping forward +1 hour at 02:08 & 02:38 respectively on the two instances.

So my question is WHY?

All Servers (PDC, Physical & VM) have the exact same Locale & Timezone Settings (GMT, Ireland(English)) All Servers have the "Automatically Adjust for Daylight Savings Time" option selected. All Non-PDC Servers are setup to use the Windows Time Service to get time updates from their PDC if there's any variances All Servers have the exact same W32Time Registry Settings

So I am completely stumped why these 2 cluster nodes decided to wait 68 minutes & 98 minutes to do their DST updates.

Thanks if anyone can shine a light on this one.

I've written a wee batch script to backup some stuff on my desktop and dump it over the network to a share... The only thing that's bugging me is the verbosity of rar.exe while it's doing the compress part of the job.

Anyone know if it's possible to tweak rar.exe to display a status bar for the entire process of at least suppress the output of each individual file being added to the archive.

Alternatively, if theres any other command line compression tools for windows (winzip/7zip) that people are familiar with, I'd be happy to use those either. Or any suggestions to make this batch job a little more robust.

@echo off
echo "Beginning Compress Job"

"C:\Program Files\WinRar\rar.exe" a -agyyyy-MM-dd -r "C:\Archives\" "C:\Work\Folder1\" "C:\Work\Folder2\" "C:\Work\Folder3\" 

echo "Files backed up to C:\Archives"

NET USE U: > %TMP%\JunkerFile 2>&1
IF %ERRORLEVEL% == 0 GOTO Success
echo "Drive Disconnected - Cannot Copy Backup To U:\"
GOTO End
:Success
        echo "Drive Connected"
        echo "Performing Backup"
        move C:\Archives\*.rar U:\Archives\
        echo "Backup Complete"
        GOTO End
:End

pause