My notes, things I have learned, and documentation I have created.

This is the process I used to upgrade an iPhone 3Gs running IOS 4.1 baseband 06.15.00 to IOS 5.0.1 while preserving the baseband and the carrier unlock.

Software installed/downloaded:

OS: MacOS Lion 10.7.2
iTunes: 10.5.2
redsn0w_mac_0.9.10b1.zip
iPhone2,1_5.0.1_9A405_Restore.ipsw

I plugged the phone into my computer and backed up from iTunes by right mouse clicking on the phone and selecting “Backup”.

I downloaded the redsn0w application and the IOS install into a folder called JB.

I unzipped the redsn0w_mac_0.9.10b1.zip and executed redsn0w.app from the redsn0w_mac_0.9.10b1 folder.

Clicked Extras.

Clicked Custom IPSW and browsed to the JB folder where I downloaded the iPhone2,1_5.0.1_9A405_Restore.ipsw from Apple.

Since I had an original 3Gs, I selected “No” on the “Will this custom IPSW be used on a newer (fixed) version of the iPhone3GS?”.

This created a custom NO_BB_OLDROM_iPhone2,1_5.0.1_9A405_Restore.ipsw file that will not attempt to update the baseband when IOS 5.0.1 is restored.

Once created, I needed to put the phone in DFU mode. For some reason, I had trouble using the method I had used before. This is what I did this time:

With device powered on,unlocked and connected via the USB cable with iTunes running:
Hold the Power and Home buttons.
When the screen goes black, count to 2.
Then, release the Power button, and hold the Home button for 10 – 15 seconds.
iTunes came up and displayed that it was in recover mode. I tried restoring my custom IPSW from here but I kept getting a 1600 error. So, once I had in recover mode, I fired up Redsn0w again and went into Extras and clicked on “Pwned DFU”.
This put it in DFU mode quickly.

Then, I was able to hold “alt/option” and click the “Restore” button in iTunes.

This allowed me to browse to the custom IPSW: NO_BB_OLDROM_iPhone2,1_5.0.1_9A405_Restore.ipsw.

This installed IOS 5.0.1 on the phone and preserved the baseband (06.15.00) with the applications and settings restored.

Once up, the phone still needed to be jailbroken and unlocked.

I closed iTunes, and brought up Redsn0w again. I needed to point redsn0w to the unchanged iPhone2,1_5.0.1_9A405_Restore.ipsw file. This time, I chose Jailbreak. The only option that I had selected was “Install Cydia”.

After rebooting and verifying the Cydia icon is installed, I needed to unlock the phone.

I launched Cydia and clicked on Manage and then Sources.

Then, I added the Ultrasn0w repository by clicking “Edit” and “Add”. Then, I added the repository: http://repo666.ultrasn0w.com. Note: this took me quite a few attempts on the 27th, due to server load some where along the line. Eventually, I was able to get the repository added.

Once added, I was able to search for and install Ultrasn0w.

I rebooted and the phone was updated, jailbroken and unlocked.

I needed to migrate a vCenter server that I had installed as a VM in a vSphere 5.0 environment. I wanted to get it moved to a different datastore. I had all the ESXi servers under the vCenter control. I needed to migrate it like this to get it to high availability (HA) storage. This is what I did to move it. The key was the first step. I am sure there are several other ways migrate the vCenter VM, especially if your datastores are already on HA storage.

From the vSphere client logged into your vCenter server, disconnect the ESXi host that your vCenter server is running on, as well as the one you want to migrate it to.

This next step could be done before or after disconnecting the hosts from vCenter:
Login to the vCenter server and install the VMware Converter.

Launch the converter and set up a job to migrate the vCenter VM to the new host location. I needed to migrate to new host, because that is where my HA storage was located.

Once the converter is completed, I shut down the old vCenter server, and started up the server in the new location.

The network did not migrate properly for me. I don’t know if this is something I missed or just the way the converter works, but I needed re-configure the IP address. That was not preserved in the migration. Perhaps, because of the change in the MAC address. Once the IP address is changed, you can either reboot the vCenter server or re-register the name with updated IP address. For me, the DNS record was updated, because the converted vCenter server picked up a DHCP address instead of the static I had assigned the server before the migration.

After I got the IP/DNS issue resolved, I connected to the vCenter server via the vSphere client like I did before the migration. Then, I just neeed to re-connect the ESXi hosts and everything was back up and ready to go.

Note: No other guests needed to be shutdown during this process.

More old notes:

These are the steps I used to create a one way trust between a Windows 2008 R2 server domain and a Windows 2003 server domain. The object was to give the Windows 2008 domain environment (DomainA.com) access to the Windows 2003 domain environment (DomainB.com), but DomainB.com would have no access to the DomainA.com domain. This process can be very confusing and difficult to keep straight in your head, but I am certain that these steps worked in the environments I described above.

Windows 2008 –> Windows 2003.
From DomainA.com (Windows 2008):
Bring up “Active Directory Domains and Trusts”
Right mouse click on DomainB.com and go to Properties.
Click on the Trusts tab.
Click “New Trust”
Next
Trust Name: DomainB.com
Forest trust
One-way: incoming
This domain only
Trust Password
Next
Next
No,do not confirm the incoming trust.

From DomainB.com (Windows 2003):
Bring up “Active Directory Domains and Trusts”
Right mouse click on DomainB.com and go to Properties.
Click on the Trusts tab.
New Trust
Next
Trust Type: Forest trust
One-way: outgoing
Sides of Trust: This domain only
Forest-wide authentication
Trust Password
Next
NExt
Confirm Outgoing Trust: Yes, confirm the outgoing trust.

Then, to grant authentication permission:
From the Active Directory Users and Computers on the DomainB.com server:
Click View and Advanced Features.
Right click on Domain Controllers and go to Properties.
Then click on the Security tab.
Click Add…
Click Locations…
Select the DomainA.com from the list and click Ok.
Then enter under “Enter the object names to select” the user/group that you want to grant access to DomainB.com from a DomainA.com account. In my case, I just chose Domain Users.
From here I am prompted for DomainA.com credentials which I enter. This can be an issue that I need figure out, because what if I have no credentials in DomainA.com? It does seems that I shouldn’t need any in this configuration. In my case, this allowed a DomainA.com login the capability to login to DomainB.com computer and access resources on DomainA.com and DomainB.com as permitted. However, because it is a one way trust, DomainB.com cannot access resources on DomainA.com.

I am getting caught up on some old notes I had not finished up. The following is the procedure I used to add a second domain controller and DNS to an existing domain with one domain controller and one DNS. This environment was all Windows 2008. I am not sure whether it was R2 or not, but I the procedure should be very similar for either version.

From the newly created member server:

Go to “Server Manager” and click on “Roles.”

Click on “Add Roles.”

Select “Active Directory Domain Services.” This will install the domain services. Since this is a second domain controller being added, you will not be asked to install a DNS server like you would if it was the first domain controller in the domain.

Once completed, you need to run dcpromo.exe to make your server an active server as an active domain controller in your domain. I chose to “Use advanced mode installation.”

You will need to choose your deployment configuration (Existing forest or new domain in a new forest). Click “Existing forest” and “Add a domain controller to an existing domain.”

Enter the name of the domain. Note: this will already be populated correctly, if you had previously joined the domain as a server.

Under “Alternate credentials”, click “Set…” and enter Administrator level account credentials for the domain and click “Next.”

Under “Additional Domain Controller Options”, I unchecked “Global catalog” but left “DNS server” checked because I wanted to add another DNS to the domain. I did not want a “Read-only domain controller (RODC).”

If you receive the following message, you need to choose appropriately. I decided to “Transfer the infrastructure master role to this domain controller.”

For some reason, I got the following message regarding my IP address assignment. I only had one NIC, and while the interface was using DHCP, I configured the interface with a static address before I started this process.

I decided to chose “No, I will assign static IP addresses to all physical adapters.” This took me backup to dcpromo wizard, which I promptly canceled to check my configuration.

I verified my IP address configuration and rebooted. The same thing happened. I went on with dynamic option, then got an message regarding DNS delegation.

It turns out that I had couple issues to deal with. I needed to disable IPv6, and I had a reference to an old secondary DNS. I removed old secondary by right mouse clicking on the domain name in “DNS Manager” on the primary DNS, and going to properties. From here, I chose “Name Servers” and removed the old DNS.

I started this awhile ago and I just want to get the information posted for my reference. I had a Windows 2008 x32 (DCOLD) server acting as my primary domain controller and DNS. I wanted to introduce a new Windows 2008 R2 x64 server and promote it to take over all the roles of DCOLD.

Windows 2008 DC (DCOLD)

Install Windows 2008 R2 x64 (DCNEW).
Ensure that network settings are configured to support DCOLD (Use DCOLD as DNS, make sure network domain name is set to DCOLD domain name).
Join DCOLD domain as a member server.

DCOLD (Windows 2008 32bit):
Copy the support\adprep directory from the installation source iso.
Run adprep.exe /forestprep

This did not work because I initially copied adprep from the 32bit source.
I got the following message”

Forest-wide information has already been updated.
[Status/Consequence]
Adprep did not attempt to rerun this operation.

Solution (You need to use adprep from the 64bit installation source.):
I had to copy the support/adprep directory from the Windows 2008 R2 CD, and run the following

adprep32.exe /forestprep

Install Windows 2008R2
Configure name/network, and bring into domain.
Run dcpromo.

Transfer roles 5 roles to DCNEW:

Schema Master
RID Master
Domain Naming Master
Infrastructure Role
PDC Emulator

This is a very basic outline, and there are several clean up steps that need to occur after this if I wanted to demote and remove DCOLD.

I found myself having to restore a Microsoft SQL 2000 database using only a file system backup. I was very cautious about my approach, because I wanted to restore the database to a database with a different name. I did this by using “Attach database …” from Enterprise Manager.

Right mouse click on “Databases”.
Go to “All Tasks”
Then, “Attach Database…”
Browse to MDF file that you restored (MDF and LDF should be restored from file system backup), by clicking the “…” next to the “Verify” button.
Then, database name you want to use for the restored database in the “Attach as” field.
Then, choose the database owner in the “Specify database owner” drop down.

It worked great, and was simple. Of course, I know that I should be doing the restores from a SQL backup, but circumstances came up where that was the only backup I had to work from. I will probably verify this procedure on a more current version of SQL server and update accordingly.

I decided to go from one extreme to the other. I installed SQL Server 2012 RCO to try to restore a database from a file system backup. It is a little different, but basically the same steps.

Right mouse click on “Databases” and click “Attach…”
Click the “Add…” button and add the mdf file you restored.
Then double click the name under the “Attach As” column heading to restore to a database with a different name.
Click “Ok” to finish.

mount.cifs has been moved from the samba-client package, for recent Redhat variants, to cifs-utils. I have been stung by this a couple times, so I figured I would note it here.

I noticed that some of my servers were running a different kernel version even after I ran yum update. I thought maybe I had not installed the centos-release-cr-6-0 rpm on all of them, but I did. I just needed to clean out the yum cache using the following:

yum clean all

I have found this very helpful when dealing with external drives to use in linux, MacOS or Windows:

mkfs.vfat -F32 /dev/sde1