[Editor's Note: Share your NT discoveries, comments, problems, solutions, and experiences with products and reach out to other Windows NT Magazine readers (including Microsoft). Email your contributions (400 words or less) to r2r@winntmag.com. Please include your phone number. We edit submissions for style, grammar, and length. If we print your submission, you'll get $100.]
Disk Spanning Revisited
Regarding Brian Kowald's "Disk Spanning" (Reader to Reader, February 1999) about using the Xcopy command to copy a group of files onto several 3.5" disks, sometimes the simplest solution is also the best solution. I recommend using two DOS commands: attrib.exe and xcopy.exe. In the directory you want to copy, type
attrib *.* +a
This command sets the archive bit on each file in the directory. Then, type
xcopy *.* a:\ /m
The /m switch copies all files with the archive bit set and turns off the bit after the file copies. When the disk runs out of space, you receive the message Error copying file filename. There is not enough space on the disk. You simply insert another disk, press F3, and press Enter.
—Andrew J. Milbrath
amil@mcw.edu
Selecting Groups of Files
You probably know how to select multiple objects in a window. For example, to select all the files in a Windows Explorer window, press Ctrl+A; to select a group of files, click the left mouse button at the beginning of the list, press Shift, and click the left mouse button at the end of the list; to select miscellaneous files, press Ctrl and click the left mouse button on the files you want to select. But you might not know how to select multiple groups of files in a directory. To accomplish this task, click the left mouse button at the beginning of the first group of files, press Shift, and click the left mouse button at the end of the first group; then press Ctrl, click the left mouse button at the beginning of the next group of files, press Ctrl+Shift, and click the left mouse button at the end of the group.
—Fernando Pessoa Sousa
fpsousa@ingr.com
USB Support
My company is planning to upgrade its PCs to Windows NT 4.0. However, several of the computers have Universal Serial Bus (USB) ports, as do most new systems. NT 4.0 doesn't support USB, which causes a problem with accessories such as printers and plotters. I'm wondering whether anyone offers a USB support patch for NT 4.0, or whether we're stuck waiting for Windows 2000 (Win2K).
—T. Talboys
ttalboys@epix.net
Cloning an NT Server PDC or BDC
In the November 1998 Tricks & Traps, a reader wanted to know how to clone a PDC or BDC to upgrade software and hardware without causing downtime, and how to change the name of the domain controller and retain its status as a PDC or BDC. A coworker and I recently had an opportunity to attempt this task. We upgraded a 133MHz Pentium server with a 60GB SCSI RAID drive to a 350MHz Pentium II server with two 60GB SCSI RAID drives in volume. We cloned the server, renamed it, and connected it to the LAN. Our downtime was less than 15 minutes for the physical move and 4 hours for the volume creation. We could have avoided any downtime if we'd had two new RAID drives to use.
The server was a Windows NT 3.51 machine with Service Pack 5 (SP5) installed. It acted as a file server with FTP, and it spooled seven printers (mostly HP LaserJet and IBM Lexmark printers). We wanted to completely update the server's hardware and upgrade the software to NT 4.0 SP4, while retaining the accounts, access, shares, FTP, and printing resources.
Before performing the upgrade, we needed a suitable backup of the OS. We made sure the NT 3.51 server used NT's standard VGA driver and that the driver was set to 16 colors at 640 * 480 resolution, because NT always starts with the standard driver regardless of the video adapter you're using. (Before we checked the server's driver, we encountered the blue screen of death several times on startup.) We also made sure that pcANYWHERE was uninstalled, because this program prevents the server from starting. Because the upgraded server would have a different LAN adapter, we added the server's driver to the live server. Thus, the server would recognize the new card after the restore. If we'd had to use a different SCSI controller, we also would have had to install the SCSI controller's driver on the live server for the new server to start properly after the restore. We used Backup Exec to create two full backups of the OS and the Registry, and we created two Emergency Repair Disks (ERDs).
On a workbench, we prepared the new server with its LAN card, SCSI RAID controller, and new 60GB RAID drives. The RAID controller took several hours to perform a low-level format. We then installed NT 3.51 on the new server. We partitioned the RAID drive with 2GB for the OS and 58GB for the files that were a volume on the live NT 3.51 server's existing RAID drive. We used the same server and domain names on the new server as on the live server. We also used the same IP address because our company's servers have a static IP address.
Next, we performed a restore from one of the backups we made of the live server. When we rebooted, we received driver errors regarding the old LAN adapter and the VGA card. Thus, we opened Control Panel and disabled the LAN adapter and VGA card. We checked to ensure that all the accounts, access, shares, and printers were still available. We backed up our work, upgraded the OS to NT 4.0, and installed the Option Pack to enable FTP. This process went smoothly, and we then made another backup and applied SP4.
We renamed the server and changed the IP address so that we could connect the server to the LAN. We had to reboot before the changes took effect. We hooked up one PC and the server on a hub to start. The server retained its status as a PDC, and we were able to log on.
The real test came when we connected the server to the live LAN. We shut the server down, connected it to the live LAN, and powered it up. We were impressed that the new server became the BDC, and both servers continued to run all the resources. We configured and tested the new FTP service, and we tested all the printers. We had to reinstall only the Lexmark printers. We were happy with our results so far.
Our next task was to copy all the files and directories from the old server to the new server. We used the Microsoft Windows NT Server 4.0 Resource Kit utility Scopy, which gave the new server's hard drive access rights to the files and directories. The copy process took almost 24 hours, so we ran it after hours to avoid downtime. When the process completed, we were ready to move the old server's RAID drives to the new server.
When we were ready to perform the upgrade, we powered down the old server, disconnected it from the LAN, renamed the new server with the old server's name, and changed the new server's IP address to match the old server's address. We shut down and restarted the new server, and the server rebooted as the PDC, with all the resources running. We rebooted the old server and performed a differential backup and restore to refresh the new server's files.
We would have been finished at this point if we hadn't had to add the RAID, but we needed more disk space. We performed a low-level format on the old server's RAID drive so we could port to the new server and volume. We shut down the new server and moved the RAID drive to it. We were then finished with the old server, so we shut it down for the last time. We connected all the cables on the new server and started the server with the two 60GB RAID drives from the old server. The RAID drive appeared in the Disk Administrator and was ready to become the volume.
We waited a few weeks before creating the volume to be sure the server was stable and so that we could make a few backups. Finally, we created a 118GB volume. We had to wait 4 hours after reboot before the server was available.
—Robin Pelletier
rpelletier@stssystems.com
Prev. page
[1]
2
next page 
