Management, in all its
forms"> eWeek: Ed, whats FN Manufacturings big network radar blip right now? Benincasa: On the application side, for the stuff thats running on the network, were going to be putting in an [enterprise resource planning] upgrade. Were also looking at online document management, but its not just one facility. For engineering documents, its a worldwide project. That will, obviously, put some demands on the network, the capabilities, the LAN links and storage requirements. The other issue that were trying to resolve is, like everybody else, security in patches. They are a constant issue for us, and were struggling with them every day to resolve. Were also looking at wireless but havent gotten that heavy into it yet.eWeek: In terms of being able to bill more accurately across business units? Gunnerson: In order to drive cost down. What weve been looking at is methods of consolidating--bringing groups together where they might not have been together before--regionalization where it makes sense, splitting out the cost of a circuit and then figuring out ways that we can work within the system to make it cost less. eWeek: Have any of you made a decision as to whether youre going to try to handle things such as patch management of your PC and other rich clients by a push technology, where the network is the means by which you get patches, security updates and other things out to the client devices? Benincasa: Were looking it at. Originally, we were running [Microsoft Corp.s] SMS [System Management Server], but were pretty much shutting that off because it just didnt perform well for us. So were back at square one, looking for something to use. eWeek: The performance problem being what? Benincasa: There were all kinds of issues with SMS. As a matter of fact, the server right now is just off in oblivion--it has to be completely reinstalled again. Were relooking at other options, but we have not come up with a list at this point. We do need something automated, and its not just the patch management; its also, how do we successfully test all of this, especially on application servers, before we roll them out? Weve got a lot of issues, and I imagine everybody else does. eWeek: It sounds as if one of the key problems for all of you is the need for a stable, reliable 24-by-7 client network while at the same time being constantly in a position of needing to evaluate and test and install changes driven in large part by the security concerns. Nowicke: Were planning to set up an independent network so that we can test some of that stuff. Even though we have a pretty heavy staff load to help do that, you reach a point where there are only so many hours in a day. It gets to the point where, do you have time that you can sacrifice to do the testing and the patching, or do you need to really rob from that activity and go back to doing the maintenance and the setup and all of the other stuff? At the same time, youre constantly trying to keep pace with disk requirements and memory requirements and user requirements and supporting users. It just gets to the point where you kind of hit your head on the ceiling.
Gunnerson: Were looking at lots of different things [at Gannett], but one of those is to try to look at the complete cost factors involved in any circuit, and were trying to change the rules.