cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
7937
Views
5
Helpful
5
Replies

Best Practice - Replacement Plan for Routers and Switches

lbrownlee
Level 1
Level 1

What is the best lifetime replacment plan?  End of Life, End of Service, # of years in service?

Please advise.

5 Replies 5

Jon Marshall
Hall of Fame
Hall of Fame

lbrownlee wrote:

What is the best lifetime replacment plan?  End of Life, End of Service, # of years in service?

Please advise.

Very debatable question.

EoS = End of Sale and not End of Service and to some extent this can be the deciding issue eg. you need to buy a new L3 switch and the ones you currently have are now EoS so you can't buy them anymore, at least not from Cisco. So this often forces your hand.

EoL = well in terms of switches/routers it's quite common to have upgraded/replaced kit before it ever reaches EoL status. Remember there is a quite a gap between EoS and EoL. Having said that in any medium to large enterprise environment i would be looking to replace EoL kit if it hasn't already been replaced for other reasons.

Other factors that determine upgrading kit -

1) new features that are not available and are not going to made available on the kit you have but that you need for your business

2) increased throughput needs. Yes a network should always be designed with future proofing in mind but what may have seemed adequate 2 years ago eg. 1Gbps linecards in your 6500 switch may now be a limiting factor and you need 10Gbps capable linecards in your 6500. Bear in mind replacement does not always mean the chassis and indeed with modular switches for example you are paying for the ability to replace and upgrade individual components without replacing the entire chassis.

So you can't and shouldn't really just pick one ie. EoL/EoS etc. as this is too inflexible an approach to take.

Jon

Leo Laohoo
Hall of Fame
Hall of Fame

I'm with Jon here.  EoS/EoL has got nothing to do with it.  If the need arises for you to buy switches with newer capabilities then that should be the case.  Of course, financial capability plays a major important role too.

Take our situation, for instance, the reason why we have started purchasing more of the 2960S is because of it's stacking capability.  We only deploy plain 2960 if the situation arises (like the project happens to annoy us).  However, the reason for getting the 3750X over the 3750 is simply "bragging rights".  He he he ...

lbrownlee
Level 1
Level 1

If you work for management, they want a plan for replacements and it cannot be on a technology not yet implemented because those are special plans and special budgets.  There needs to be a replacement plan in place for allocating budgets.  Does anyone have a plan in place for replacing equipment outside of processing needs?

lbrownlee wrote:

If you work for management, they want a plan for replacements and it cannot be on a technology not yet implemented because those are special plans and special budgets.  There needs to be a replacement plan in place for allocating budgets.  Does anyone have a plan in place for replacing equipment outside of processing needs?

Well then the EoS/EoL deadlines can be your starting point. The point i was trying to make is that often you don't reach those milestones.

But as a network engineer/designer you also need to be aware of what features are coming down the line into which products and what the trends in terms of bandwidth etc. so you can steer management in the right direction and also incidentally protect your own job eg. management won't be too impressed if you pick gigabit switches and 1 year down the line you then need 10Gbps connectivity in some areas.

EoL is the most obvious one, you shouldn't have switches that are in your network that are EoL because if things go wrong/bugs develop etc. you are in your own. If nothing else replace switches/routers that are approaching EoL in your estate.

EoS is not so important.

Last place i worked as a network designer i first identified switches which were fast approaching EoL. EoS didn't affect that decision other than obviously i couldn't purchase these and wouldn't want to anyway as EoS then becomes EoL.

I then looked at developing trends within the networking field, sat down with the Unix/Windows server people, desktop people etc. to see what technologies they were introducing and what bandwidth requirements they had and take that into account when specifying the new equipment.

Finally sometimes it is just up to you to make a decision and then justify it to management. When Cisco initially released the service modules within the 6500 chassis we saw this as an ideal opportunity to quickly deploy new services for the server guys being able to offering firewall, load-balancing/IDS if needed without having to visit the DC each time for physical cabling etc. We argued the case and management accepted it.

So it's still a combination but if you do nothing else replace EoL.

Edit - by the way, not sure what you mean by "work for management". We all have to report to management, although often many of us probably wish we didn't

Jon

Cisco generally sale a product for 5 years and support it for ano

ther 5 years after sale. What they dont do is publish when the product first came on the market

, so you could then decide your own pathway.

Getting Started

Find answers to your questions by entering keywords or phrases in the Search bar above. New here? Use these resources to familiarize yourself with the community: