Info Systems will need to upgrade the operating system for our main network router/switch sometime during the next several weeks. The information regarding this planned outage includes the following:

PROPOSED LENGTH OF TIME: about 3 hours.

SYSTEMS AFFECTED: The ENTIRE EMU computer network. Network connectivity for all desktop computers, servers, printers connected to the EMU network and the Internet will NOT function during the outage. (This is MAJOR network “brain surgery”).

All incoming and outgoing email will be queued during the outage (none should be lost). Anyone on the Internet attempting to connect to any of our web servers (www.emu.edu, http://blackboard.emu.edu, http://mail.emu.edu) will get a “cannot connect to the server” error message. Windows computers connected to the Novell servers will be warned about 10 minutes before the outage that they should save any open documents and logout. We will also send regular email reminders leading up to the outage so everyone is aware of the time that the outage begins.
———————–
Upgrade SUCCESSFUL at 5pm, Fri, 7/25, one hour AHEAD of schedule.

Kudos to Dan Marple Jr of Info Systems and Mike Durkin of STG Networks of Winchester!

A drive is failing on server NS.

I called Dell and another one is being sent. I checked the drive using Dell’s raid disk test utility. It failed the “smart check”.

I took the “spare” 36 gig drive from MS2 and installed it into the system. The replacement drive will become our spare.

Dan Marple Jr.

I disabled sophos on FS. On the corrupted file I found the following:
couldn’t check properties in explorer
couldn’t copy the file in explorer
couldn’t delete the file in explorer
couldn’t copy the file in a dos box
COULD delete the file in a dos box.
I then rebooted the server just to make sure it was happy.
It appears that if there is a high CPU utilization we should check what process is the highest. If it is sweep, check the last file it scanned – we might be able to use that to track down a corrupted file (if that is the problem). In this case “user” was the last user to have a file scanned, but it wasn’t the corrupted file.

I was hoping to be able to get a copy of the file and then place that on a test server to see if we could duplicate the error and see if there was another way to prevent the server from crashing, by disabling sweep w/o a reboot.

Servers NS and STAPPS were rebooted due to an error when running dsrepair. NS’s problems was not able to find the files – in which rebooting was the easiers fix. STAPPS’s problem was it hung while running DSREPAIR. All network access was still there, but backups would not have occured.

FS server crashed due to overloading of CPU. Why – yea – good question.
Attached sniffer to look for problems – maybe we will catch something if it happens again!

Had to restart, run vrepair, and it was good to go.

A drive failed in server STAPPS – one of the 2 4G sys drives.

The drive was reseated and it rebuilt.

The server FS and FSAPPS was rebooted to resolve some locked file and locked user issues. It appears to have worked. While the servers were down images were also made.

FS did not go down gracefully. To bring it down I had to power it off. There were several users with files open – were they “really” open, or were they in an ambiguous locked state similar to the files we were trying to free up? Because of this I had to run vrepair on all volumes – there were many errors on userdata. After the server was back up, there were not any problems with the files we were trying to “free”.

Upgrade the 6509 IOS and the MSFC IOS to allow for the use of the new copper based GBICs. Changed all trunking from ISL to DOT1Q. Had problems in the seminary with the link between the 3550-48 and the 3524 on the 1st floor. It is possibly a bad gigastack on the 3524.

Upgraded the Pix to IOS 6.2.2 to allow for testing and use of VPN’s for Lancaster as well as for the windows client.

A flash card will now need to be purchased for the 6509 to allow for failover and for future upgrades – not enough memory on the board anymore.

All Novell servers, but NS, are patched to the latest service packs for pervasive software. I’m leary on doing NS since these SP’s affect Btrieve, which Novastor uses.

Dan