FW: [IUG] Mouse activity and Millennium freeze-ups

["Hennessey, Christina" <CHennessey at lmu dot edu>]

Hi folks,

I wanted to follow up on my message from November 30, 2005 (included below) in case another library runs into similar problems.

There is this item in the CSDirect "Access through Firewalls" FAQ (December 6, 2005 version):
http://csdirect.iii.com/faq/firewall.shtml 

**
Are there any known issues related to running the Millennium clients through a Cisco PIX firewall?

Yes. Cisco PIX firewalls have the ability to alter certain connections as they traverse the firewall. Cisco PIX refers to this as a "fixup," which can be enabled or disabled for several network services including the Skinny Client Control Protocol (SCCP/skinny). Unfortunately both SCCP and Millennium use port 2000/TCP. When a PIX firewall sees traffic on port 2000/TCP it assumes it is a SCCP connection and attempts to alter the traffic. This creates problems with Millennium connections.

Innovative recommends that you or your firewall administrator configure your Cisco PIX to disable SCCP fixup using the command "no fixup protocol skinny 2000" on the firewall.

**
Once our university IT disabled SCCP fixup on the firewall, we immediately had no more problems with Millennium freeze-ups. 

Christina Hennessey
Systems Librarian
Loyola Marymount University
chennessey at lmu dot edu


-----Original Message-----
From: Hennessey, Christina 
Sent: Wednesday, November 30, 2005 5:49 PM
To: 'IUG INNOPAC List'
Subject: RE: [IUG] Mouse activity and Millennium freeze-ups

Mark,

I was actually planning on posting a similar query to the listserv today, so it's great (in a way) to see that we aren't the only site with this problem. 

We have been working on a similar problem at our site with III for a few weeks with no cause identified as of yet. We also see it across all our modules, and even on our newest systems that have 1 GB of RAM. Almost all our systems are Dells. The problem does not seem related to networking, any particular function within a module, or amount of time the module has been active. We are still running Silver, with a new server we installed in March of this year. We're just a single, standalone site so the amount of staff using the system is minimal.

III has restarted several processes on our server a few different times, and the MilData port, but that hasn't seemed to help. Our application freezing is "true" freezing, where the application turns completely gray and becomes unusable (and not just a network timeout).

As you are experiencing, this is creating quite a negative attitude towards the Millennium modules amongst our staff, as we are also trying to aggressively move staff to Millennium modules. We started seeing these problems as soon as more staff moved to Millennium, although about 25% of our staff were already using Millennium modules before this push.

Any more tips the III world could offer us would really help.

Thanks,
Christina Hennessey
Systems Librarian
Loyola Marymount University
chennessey at lmu dot edu

-----Original Message-----
From: Mark Stoffan [mailto:mstoffan at wncln dot lib dot unca dot edu] 
Sent: Tuesday, November 29, 2005 12:10 PM
To: 'IUG INNOPAC List'
Subject: [IUG] Mouse activity and Millennium freeze-ups

We're a consortium of three libraries sharing a single Millennium
system. For the past few months, we've been aggressively trying to
migrate our staffs away from the Telnet functions by encouraging use of
the Millennium modules as much as possible. Many staff members report
frequent client freezes, as many as 9-12 per day, which is generating a
negative mind-set toward use of the Millennium clients. In speaking with
help desk staff, they recommended the usual actions such as documenting
the specific activities of the staff members, identifying conflicting
software, etc. We've been doing this now for some weeks.

* The problem occurs in about 75% of the clients.

* It occurs at all three sites.

* It occurs in all modules.

* There is no particular workflow, procedure or function identifiable as
a common thread.

* Many of the machines experiencing the problem have no software
differences from the machines that don't experience the problem.

* Network issues don't seem to be a factor.

* The computers across the three campuses are a mix of Dell and Gateway
PC's of various models. All are Pentiums having at least 512 megs of RAM
with many having a full gigabyte, and they meet or exceed the
requirements listed by III. 

* The Java engines have been updated to the current release on CSDirect.

* We're running Release 2005 LE (scheduled to install 2005 and 2006 LE
in early January)

The only thing we've noted is that users who rely on hot keys rather
than the mouse to move around the screen experience a marked reduction
in freeze-ups. Several staff members have pointedly avoided using the
mouse and the their incidences of lockups dropped from several per day
to only 1-2 per week. 

Has anyone else experienced this type of behavior and can any additional
light be shed on it? Some staff estimate they lose 10% of their output
so it's not just a minor, occasional annoyance. I'd really like to
figure out what's causing this problem and solve it if possible.

Thanks,
Mark

+++++++++++++++++++++++++++++++++ 
Mark A. Stoffan 
Network Librarian, Western North Carolina Library Network 
1 University Heights, CPO 1500 
Asheville, NC  28804                        (828) 232-5095 (office) 
mstoffan at wncln dot lib dot unca dot edu        (828) 231-6285 (cell) 
wncln.lib.unca.edu                            (828) 232-5137 (fax)