Renewing an Expired SBS 2003 SSL Certificate

Most of you have probably already renewed your certificates as they started to expire in 2008.

Its a simple procedure it involves running the "Connect to the Internet" wizard in server configuration.

To renew an sbs2003 Certificate:

1. Click Start, and then click Server Management. In the console tree, click Internet and E-mail. In the details pane, click Connect to the Internet.
2. On the Connection Type page, select Do not change connection type, and then click Next.
3. On the Firewall page, select Do not change firewall configuration, and then click Next.
4. On the Services Configuration page, ensure that the appropriate network services are selected, and then click Next.
5. On the Web Services Configuration page, ensure that the appropriate Web services are selected, and then click Next.
6. On the Web Server Certificate page, select Create a new Web server certificate, type the fully qualified domain name (FQDN) of your server (for example: server.contoso.com), and then click Next.
7. On the Internet E-mail page, select Do not change Internet e-mail configuration, and then click Next.
8. On the Completing the Configure E-mail and Internet Connection Wizard page, click Finish.

Now this is all good if it succeeds.

If you get an Error on the FIREWALL configuration stage click OK to continue, now you will need to restart the server to be able to restart the HTTP SSL Service which when you try to start it an error occurs stating Error 2: the File cannot be found.

After restarting our server this error disappeard and we had a new SSL certificate for another 5 years.

-Fr33ze

Exchange and IIS - Why these two should be sent in either direction.

So i had an issue with Mail not senging out after i removed an ISA server from the network.  Oh, i thought to myself i havent removed the SMARTHOST from the SMTP Connector.  Sure enough they it was...so removed it and selected DNS to send email.

Waited.....and waited.....and waited....and...... well the email was still not sending...GRRRRRRR

Checked all avenues the email wasnt even trying to send through the new firewall it didnt even know it existed yet there was alot of traffic hitting the firewall but none of it SMTP.   hmmmmm

I increase the Diganostic logging level on the MSExchangeTransport and found that my email server was still trying to send to the SMARTHOST, but WTF the smarthost is nowhere to be found in the system....or IS IT!!!!.

Ladies and gentlemen we have a problem with Exchange updating the IIS Metabase.

Using the Matabase Explorer found in the IIS 6.0 Resrouce Kit i was able to find the the SMTPSVC/1/DOMAIN had a few references in it that included the Default SMTP * also domains that were hosted by the Exchange server AND the bloody SMARTHOST was still listed in here.



Checking the data of the SMARTHOST entry i noticed that the routeaction value of the * domain was 2 and the routeaction of the SMARTHOST was 0, changing the SMARTHOST to 2 amd changing the * to 0 and deleting the content to the routeactionstring from the * domain freed the email to be sent over the internet rather than to a smarthost.

Although this is just a workaround the REAL problem lies with Exchange not able to update the metabase, stay turned i hope to have an answer for this soon.  If not Uninstall IIS reinstall IIS and the reinstall exchange.....OHDEAR!

-Fr33ze

NETBIOS NT 4.0 requires that you are still around.

So had another issue with a system that had a few NT 4.0 workstations and some Photocopiers that used SMB to scan files to the server.  One day it was working the next it wasn't.  So to my amazment (not) i discovered that someone HAD to have been playign around because this resolution to this wasnt as far out as you might expect.

NT 4.0 workstaions couldnt get onto the domain abut they could PING the domain controllers but they couldnt browse the file contents.  But for some strange reason they could browse files on other servers just not the DC.  The photocopiers used SMB to copy files to the server and that wasn't working.

I thought well since these 2 things (NT4 andSMB) where old technology there must be something OLD missing from the DC.

They NEVER had WINS installed on the DC so it cant be that.......OH LOOK HERE in the advanced section of the TCP/IP properties of the network card, under the WINS tab NETBIOS had been turned off once i turned it on again shit started to work.

Now if you have older systems like these i would suggest turning on WINS or if you have older systems that are on a WAN or seperated by a router turn WINS on.

Antoher crisis averted.

-Fr33ze

Exchange ActiveSync 0x85010014 and OWA Loading.... Screen

This problem has me going for HOURS!!!!! fucking Arhhhhhhhhhhhhhhhhhhhhhhhhhh.

After my upgrade of disks in an array to larger ones there was an issue that i noticed but though it was because the exchange databases were offline so i thought nothing of it.  Tunrs out that somewhere down the track someone had made a change which caused Activesync and OWA to stop functioning or to give strange messages.

Turns out that someone had changed the security on the EXCHANGE Virtual directory to BASIC only after changing it to include intergrated security it worked fine. 

What confused me was the client didnt explain that there was an issue with OWA aswell as ActiveSync so focusing on ActiveSync only caused me more problems it also didnt help that the event log had no leads to point to the issue.

Let me just state that if Exchange is working and you are only having Sync issues and the OWA Loading... page issue then it is purly IIS related and most likey directory security related.

Do not think that you need to reinstall exchagne or IIS or even delete the Virtual directories and recreate them, just think Directory Security.

If that fails then Google is your friend.

-Fr33ze

User Profile error Event ID 1508 on Terminal Server

I had an issue where a user profile would always be in use and the user could not login to the terminal server.  I used this program called User Profile Hive Cleanup Service.

This program clears any in use user profile and allows the user to login successfully.  This is also helpful because most times you would be required to restart the server to free the profile, this one prevents this.

-Fr33ze

Replacing Disks in a RAID Array with bigger ones

I must admit this one concerned me abit because i KNEW i would have to rely on a backup to get the data back on these disks.  What was more was this array had Exchange Databases on there zzZZzz.

So a client requires us to install larger disks into an array, now the current setup is a RAID 5 array with 5 146Gb disks, we need to replace those 146 Gb Disks with 300Gb disks.

The thing is, is we have to remove the current disks and array and create a new one, practically deleting the disk from the system.  The GOOD thing is with HP these days is that a copy of the RAID Array information is kept on the disks so you can remove it from the Controller and if in the event that something goes sideways you can reinstall the disks IN THE SAME ORDER THEY CAME OUT and get that array backup and running.  So when removing diskd from and array keep them in the order that they came out of the cage if you dont you will lose the array configuration.

So i started my job by doing my OWN backup using Shadow Protect i also knew that the client was doing a backup of the system using ARC SERVE to tape.  This server had 2 Arrays one mirrored set for the OS and a RAID 5 for the data (the RAID 5 is the one we are updating).  I backup up the ENTIRE SERVER so if in the event somethign REALLY when sideways i knew that ShadowProtect would be able to recover from it within minutes (depending on the size of the backups).

So after i had performed a backup and the client had comfirmed that their backup was successful i turned of the server and removed the 5 disks that were part of the RAID 5 array.  I installed the new disks into the server and turned it on.

I got a message syaing the the OLD array was degraded and there were members missing, it asked if i would Press F1 to continue booting or F2 to accept the lost data now here i pressed F1 because i knew that the BIOS was going to ask me to enter setup by pressing F8 so PRESS F8.

In the Array BIOS setup screen i selected CREATE ARRAY except there were no disks available, this was an oversight as the OLD array that we had removed was still in the controller, we will need to delete this array, select DELETE ARRAY and delete the array that you removed.  Once this is done you will select CREATE ARRAY and be able to select the NEW disks to create the NEW array.

I created a RAID 5 array using all the disks.

After this was created i exited the setup and let the OS boot.  Everything was going good so far, after it booted into the OS i logged in and created the Volume in Computer Management (Server 2003) i gave the Volume the SAME label and drive letter as the old one.

Next was to restore the data so to the clients tape backup we restored the data back to the drive, now depending on how much data ther is this may take some time.

After about 3 hours the data was restored, the first thing i noticed was that the shares hadn't been recreated DONT RECREATE THEM reboot the server as the shares are still in the OS they just havent been applied because you booted the server without any data on the drive that you replaced.  Rebooting the server will put the shares back.  Next thing i noticed was the client didnt backup the Exchange Databases as FILES which can be a GOOD and BAD thing.  Good because we had to do an ONLINE resotre using the ARCSERVE exchange agent BAD bacuse if that AGENT ahdnt worked in other circumstances they would have lost their exchange data (perhaps, but i had 2 other methods of recovering echange BACKUPS PEOPLE BACKUPS!!!!!!!) So after finding last nights tape and begining a restore from ARCSERVE the backup was failing Eeeeeeeek.  But yet again it was an OVERSIGHT on my part as i had turned the Information Store Service off, for and ONLINE backup you need to have the Information store service running.  Dont worry this service will run even if there is NO database file on the server WOOHOO!!!.

ARCSERVE had some funny agent options that you will have to bypass firstly it will try and restore it to a recovery store area, UNITCK THIS as we need to recover the prodction database also there are options in the agent to mark the databases to be able to be restored and dismount and remount the sotre after backup TICK THESE.

After the backup had run it replayed all the logs and all email was restored.

There are other methods to replacing disks but i think this is the ONLY SAFE WAY!, i cant stress the BACKUPS are the MOST important thing when working with servers and data, make it priority to backup everyday and backup EVERYTHING!!! when backing up exchange backup both the Files aswell as the Information Store using the appropriate agent.

GOOD LUCK replacing your disks with large ones.

P.s Initial Planning of disk storage over 5 years is the best method of maintenace you could ever do, plan it before you install it.

-Fr33ze

The OverHeating Laptop

We had an HP laptop come into work and it was overheating bad and swtiching off.  So after some testing i found that it was reaching 90 degrees Celcius and then switching off to cool down.

I removed the Keyboard and button plate and then checked the system again to see if the fan was working.

1. The Fan was working fine.

When the fan turns on it cools the Copper heatsink which in turn cools to processor, when touching the heat sink close to the fan it was cool and touching the heat sink close to the processor was Fucking HOT!. 

I removed the Fan and the heat sink then removed the processor i cleaned the transfer paste off both the processor and the heatsink then applied more and refit the Processor, heatsink and fan.  Started the laptop up and performed a CPU burnin using CPU BUNRIN and waited for it to overheat or to stay at a reasonable temperature.

I was able to view the CPU temperature using a program called CPUID this showed me that my efforts were fruitless and the CPU was still overheating.

Luckly this laptop was under warranty.  So i ordered and Mainboard, a processor and a new heatsink.

When these parts arrived i decided to try the easy parts first so i removed the old heatsink and processor and installed the new ones.  BINGO!!! this fixed my issue.

It seems that the CPU may have been overheating and unable to cool properly but replacing the heatsink i think was also a sure thing as i wasnt sure whether the old heatsink may have been warped or bent in some way prventing proper cooling of the CPU.

A least if the laptop hadn't have been under warranty it wouldnt have cost me the world for a new mainboard.

So if your laptop or even your PC is overheating:

1. Remove clean and replace your heatsink using new heat transfer paste
2. Replace the heatsink
3. Replace the CPU
4. Pray it aint anything else or its gonna start costing some.

Use SPUID to check your runnign temps my i7 is running at 65 degrees Celcius ATM.

-Fr33ze

IdeaPOS and the Hidden Pending Sale

So IdeaPOS client rings up and says that they can't do an end of shift because a message pops up and says "You can't perform and end of shift when there is a pending sale".  So onto the back office machine i go and do a pending sales report, sure enough there is a pending sale.

I go back to one of the tills and hit the pending sales button, but .... Nothing.... i also look at the table map and all the bar tabs, but nothing.

This confused me so what i had to do was click on Setup, Yes/No Options and click Misc.  Option 20 - Prevent end of shift when there is a pending sale.

Now this is ONLY temporary because if there is an on going issue with sales then it needs to be addresses.  Also if you have a multi laned site then you should be using the POS Server which keeps all the tables and pending sales centralised so that if there is a change on a table all other lanes know about it.  Also if the POS Server is unavilable then you cant use tables feature until you change the POS Server location.

So after contact with IdealPOS they said for me to check to Table Summary button, what drew them to this conclusion was that when i did a pending sale report to totals were displaying as if they were oweing.

The next day i revisted the site only to find that the pending sale had gone, so unfortunatly it had fixed itself or someone had come across the pending sale and removed it.

So these places here are where to check to find a pending sale:

1. Pending Sale Button
2. Tables
3. Bar Tabs
4. Table Summary Button on table map

There is also away to remove pending sales by deleting it from the clear data section of the utilities menu item.  Tick the box and then overwrite the other databases.

Before you do any work on a site do a BACKUP so that you can recover from anythign that may go sideways.

-Fr33ze

Expanding Arrays and Logical Drives in HP ML/DL servers using Array Management Utility

So i have to expand a 0+1 array with new disks, i have 6 disks already in there totaling about 410Gb, i need to add 2 new 146Gb disks.

Step 1

BACKUP!!!!!!!

Even the easiest jobs can go wrong sometimes.

Step 2

Install the 2 new disks into the free slots of the 8 slot drive bay (Comes with the ML350).

Step 3

Open up the Array Configuration Utility located HP Management tools in the start menu.

Step 4

Select the Array Card and the array that you are Expanding, in here you should notice the you will see the arrays that you have configured aswell as Free Space which is the total of the new disks you have.

Highlight the array you want to expand the Click on Expand Array.  This will give you a wizard, Click Begin.  From the next screen you will be asked to select the drives to add to the array, in my case i had matching 146Gb drives which after the expansion i would have 8 146Gb drives in a 0+1 Array.  So select the disks that you want to add to the array and Click NEXT.  Click Finish to Save the changes.  The controller now reconfigures the array so that it is using all the new disks.

Keep in mind that the system was up and runnign during this expansion but it was done out of hours so there was no need for me to shut the system down.

This process took me 4.5 - 5 hours to complete, so go home and wait till later.  I was able to remote in a check the systems progress.

Step 5

After you have expanded the array not much seems to change, that because you havent done anything with the free space we now need to Extending the Logical Drive.

Click on Extend Logical Drive, and click the drive you want to extend, next select the new size of the drive the is being extended and click finish to save changes.

This part didnt take very long i dont think, but its still got a counted and progress bar so keep an eye on it.

Step 6

After that process has finished we have finished using the Array Configuration Utility so we can close that.  Next thing is to Extend the Volume within windows.

Right mouse click on "computer" and click on Manage now depending on yr operating system you will need to find Disk Management.  Find the drive the you have expanded and extend in the HPACU and right click on the partition and select Extend Volume.

One thing to make note of is that windows wont allow you to extend a system volume so if you have done this to a system volume then you will need to image that drive extend it outside of windows and then reimage the drive, unless someone has a better idea?  I would LOVE to hear it because i hate exteneding system partitions, i like shit thats easy.

haha FYI heres the HP ACU User Manual

-Fr33ze

Don't Let Hard Disk space run out on your Hyper-V Host

So heres a pickle i got myself into and shoudlnt have.

We had a client who runs an ML350 as a Hyper-V host and has 3 guest machines an SBS2008, Terminal Server and an SQL Server.  So someone decided to dump a shit load of data onto a dynamic drive on one of the servers, which in turn used up all the remaining disk space on the HOST drive that the Guest VHD was sitting on.

If you have ever had a Guest VM display "Paused, Critical", this is one of those.  So arriving onsite to find that the Physical drive had 1mb of disk space left.  Since the only files on this drive where VHDs there was no imediate way of gaining space.

My breathing room came when i turned off the terminal server (the least important server) this cleared some Virtual Memory files from the drive which gave me about 3-4 gig of space to resume the paused servers and be able to successfully shut them down.  After shutting down the SQL i notice this in the Hyper V manager "Merging in progress" Whoops it seems as though someone has tried to delete a snapshot and not shut the server down to complete the merge.  I was presented with another problem, the merge was going to take all of my free disk space.

Sure enough i lost my diskspace again :(.

I found my solution by deleting a snapshot of the terminal server, lucky for me that the terminal server was on its own Physical HDD but the snapshots were on the disk with no space left, SO i got 9 gig back :).

Now at this point i though NO WAY was i trying anything else, so i started up the SQL server and resumed the SBS server and got them up and runnign with 4 gig free.

My solution to the problem was going to be adding new disks to the array in the server and expanding the size of the array, which i eventually did the next day and the server now has plenty of space to grow.

So rule of thumb ALWAYS make sure you have enough disk space to host Fixed VHDs of your important servers and leave the none important ones on other Physical HDDs.

FYI Go HERE this shows in quick steps how to compact a dynamic VHD if you have a disk space scare.

I will post after this as to the method i used to expand the array on the Server.

-Fr33ze