Time to gut the network - thoughts?



  • It's time to get this network to what @scottalanmiller would call home user level.

    Current setup:
    Building 1
    Master switch - HP 2824 (1 Gb)
    endpoint switch 1 - HP 2650 PWR (100 Mb)
    endpoint switch 2 - HP 2650 PWR (100 Mb)
    Wireless controller - Cisco 4402-25 WAP controller, 8 APs

    Building 2
    endpoint switch 3 - HP 2650 (100 Mb)
    endpoint switch 4 - HP 2650 (100 Mb)
    WAP switch - Cisco 3560
    WAP - 10 Cisco WAPs

    Building 1 has 3 pair of fiber running to building two (private fiber)
    Building 1 has IP phones only
    Building 1, most PCs connect through the phones for network access

    Building 2 has digital phones only
    Building 2 50% wireless laptops 50% wired PCs

    Proposal:
    Replace all switches with 1 HP 1920 and 4 UBNT EdgeSwitch ES-48-500W (Already have the HP, it's new so I should use it).
    Don't replace Cisco switch, those ports can go onto one of the EdgeSwitches
    Replace the Cisco WAPs/controller with UAP-AC Lites
    Use 10 Gb SFP+ Fiber adapters to link switches in building 2 to building 1

    Parts list

    4 Edgeswitch ES-48-500w
    4 10 Gb SFP+ modules
    4 TRIPP LITE Fiber Optic Mode Conditioning 6' Patch Cable LC/MC to SC 2M (N425-02M)
    20 UAP-AC Lite

    Considerations:
    Currently my HP-2824 switch is acting as a router between my VLANs. Anyone know from experience if the Edgeswitches can do this? The specs claim they can.

    Even if the Edgeswitch can do this, should I? Or should I install an EdgeRouter to route between my VLANs? I currently don't have any ACLs between VLANs. I have VLANs because of legacy thinking (heck, my phone provider is still practically demanding a VLAN for the VOIP phones).

    FYI - in Q1 of 2017, the plan is to replace the digital phones with IP phones, this is the reason for the POE switches in Building 2.

    Other than having the HP 1920 switch, nothing else is purchased or set in stone.

    Suggestions of changes? other questions, things I should consider?



  • @Dashrender said in Time to gut the network - thoughts?:

    I currently don't have any ACLs between VLANs.

    No zone policy either?



  • @Dashrender said in Time to gut the network - thoughts?:

    Anyone know from experience if the Edgeswitches can do this? The specs claim they can.

    Looks like it. But as you can see, I don't use it for that.

    0_1476830994512_route-table.png

    0_1476831005858_configured-routes.png



  • Can you really get more than 1Gb/s between the buildings? Is it worth going to 10Gb/s now?



  • @scottalanmiller said in Time to gut the network - thoughts?:

    Can you really get more than 1Gb/s between the buildings? Is it worth going to 10Gb/s now?

    A good point. None of the clients will really benefit from this.

    You're still bottlenecked at each client and at the router.



  • @stacksofplates said in Time to gut the network - thoughts?:

    @Dashrender said in Time to gut the network - thoughts?:

    I currently don't have any ACLs between VLANs.

    No zone policy either?

    I don't know what that is.



  • @stacksofplates said in Time to gut the network - thoughts?:

    @scottalanmiller said in Time to gut the network - thoughts?:

    Can you really get more than 1Gb/s between the buildings? Is it worth going to 10Gb/s now?

    A good point. None of the clients will really benefit from this.

    You're still bottlenecked at each client and at the router.

    Yeah I was wondering wondering that myself. Converting to 10 Gb dollar wise isn't that much more than 1Gb, but percentage wise is huge on the components, like 100%+.

    As mentioned it's probably just a waste of money.



  • With 10Gb tho, he can loose 2 of the 3 fiber runs and the network won't notice. Besides, it's not like 10Gb adapters for servers are very expensive from a place like xByte, right @xByteSean?



  • @travisdh1 That said, I don't know what the price difference actually is myself. For servers and backbone it might be worth the upgrade all around.



  • @travisdh1 said in Time to gut the network - thoughts?:

    @travisdh1 That said, I don't know what the price difference actually is myself. For servers and backbone it might be worth the upgrade all around.

    I'd have to get a different/additional switch to have enough 10Gb ports to cover servers. That kinda makes it not worth it.

    Assuming the use of only one of the ES-48-500w as a backbone switch, it only has two 10Gb ports.



  • Are you maxing out your 1Gb ports? That's what would push me to 10Gb.



  • (hijacking on)

    also thinking of upgrading some networking gear with the EdgeSwitch



  • @travisdh1 said in Time to gut the network - thoughts?:

    @travisdh1 That said, I don't know what the price difference actually is myself. For servers and backbone it might be worth the upgrade all around.

    Refurbs aren't bad, but new NICs are around $250-300 and if you use SFP then it's even more.



  • @Dashrender said in Time to gut the network - thoughts?:

    @stacksofplates said in Time to gut the network - thoughts?:

    @Dashrender said in Time to gut the network - thoughts?:

    I currently don't have any ACLs between VLANs.

    No zone policy either?

    I don't know what that is.

    Instead of straight ACL firewall rules.

    I guess I don't see the point in VLANs with no firewall rules.



  • If it were me, I'd bump those HP 2650's up to the Ubiquiti equivalents and just stick with 1Gig backbone, unless you're consistently pushing the 1gig links pretty hard. If the price difference between 1gig and 10gig SFPs aren't that big, then I would consider going to 10Gig.

    I'd also just leave the HP 2824 as the VLAN router since there's no real rules like that. Or get the EdgeRouter as a primary and keep the 2824 as a backup.



  • @Mike-Davis said in Time to gut the network - thoughts?:

    Are you maxing out your 1Gb ports? That's what would push me to 10Gb.

    No, not really. 90%+ of our network traffic goes over the fiber link to the internet. We have a 100/20 internet connection.



  • I put 10 Gb on the table mainly because it's actual cost over 1 Gb links was around $100 more per link and it would give me a chance to - just do it.

    But as I mentioned above, when I put my IT hat one, you know, the one where we do what's best for the company - which includes being financially responsible - well, I suppose I should just save that $200 because it's really not going to gain us anything.



  • @Dashrender said in Time to gut the network - thoughts?:

    I put 10 Gb on the table mainly because it's actual cost over 1 Gb links was around $100 more per link and it would give me a chance to - just do it.

    But as I mentioned above, when I put my IT hat one, you know, the one where we do what's best for the company - which includes being financially responsible - well, I suppose I should just save that $200 because it's really not going to gain us anything.

    How often are you pushing out images to a new desktop that would definitely be filling up your pipe if you're trying to do more than one or two at a time?



  • @Dashrender said in Time to gut the network - thoughts?:

    I put 10 Gb on the table mainly because it's actual cost over 1 Gb links was around $100 more per link and it would give me a chance to - just do it.

    But as I mentioned above, when I put my IT hat one, you know, the one where we do what's best for the company - which includes being financially responsible - well, I suppose I should just save that $200 because it's really not going to gain us anything.

    Does the fiber between those buildings support 10G? For 200$ this seems like a no brainer to me. It's not enough to cause any issues with accounting and it introduces some opportunity. I know, buy for what you need now and not what you need tomorrow but $200 just doesn't seem like much.



  • @stacksofplates said in Time to gut the network - thoughts?:

    @Dashrender said in Time to gut the network - thoughts?:

    @stacksofplates said in Time to gut the network - thoughts?:

    @Dashrender said in Time to gut the network - thoughts?:

    I currently don't have any ACLs between VLANs.

    No zone policy either?

    I don't know what that is.

    Instead of straight ACL firewall rules.

    I guess I don't see the point in VLANs with no firewall rules.

    Legacy understanding, and the belief (by the phone installation company) that VLANs would allow QOS for the IP phones.

    Of course the use of VLANs does allow for VLAN X to have a higher QOS level, but if the switch is saturated by traffic on other VLANs, I suppose the switch should give priority to the QOS ratings, but I probably have problems to fix.

    One thing I've considered to dumping the VLANs and moving to /23 or /22.



  • @JaredBusch said in Time to gut the network - thoughts?:

    @Dashrender said in Time to gut the network - thoughts?:

    I put 10 Gb on the table mainly because it's actual cost over 1 Gb links was around $100 more per link and it would give me a chance to - just do it.

    But as I mentioned above, when I put my IT hat one, you know, the one where we do what's best for the company - which includes being financially responsible - well, I suppose I should just save that $200 because it's really not going to gain us anything.

    How often are you pushing out images to a new desktop that would definitely be filling up your pipe if you're trying to do more than one or two at a time?

    Never.
    I don't use multicasting for deploying images, so each image would be it's own stream. I'm not sure what my disk through would give me much more than 1 Gb/s.

    The closest I come is when doing updates manually at a workstation - say to 1607. In that case, I'd probably be maxed at around 4 machines at once.



  • @Dashrender said in Time to gut the network - thoughts?:

    @JaredBusch said in Time to gut the network - thoughts?:

    @Dashrender said in Time to gut the network - thoughts?:

    I put 10 Gb on the table mainly because it's actual cost over 1 Gb links was around $100 more per link and it would give me a chance to - just do it.

    But as I mentioned above, when I put my IT hat one, you know, the one where we do what's best for the company - which includes being financially responsible - well, I suppose I should just save that $200 because it's really not going to gain us anything.

    How often are you pushing out images to a new desktop that would definitely be filling up your pipe if you're trying to do more than one or two at a time?

    Never.
    I don't use multicasting for deploying images, so each image would be it's own stream. I'm not sure what my disk through would give me much more than 1 Gb/s.

    The closest I come is when doing updates manually at a workstation - say to 1607. In that case, I'd probably be maxed at around 4 machines at once.

    Are the fiber links setup with Link Aggregation? If not, that could make a big difference.



  • @Dashrender said in Time to gut the network - thoughts?:

    Legacy understanding, and the belief (by the phone installation company) that VLANs would allow QOS for the IP phones.

    Should have walked them out the door the moment that you found out that they didn't know even the basic underpinnings of networking or phones. What value did they bring if they aren't aware of how either work? There are only two skills sets for VoIP to have... and thinking that VLANs do QoS indicates basically zero knowledge of either. That's super basic stuff. It doesn't require any special phone knowledge to know why that's impossible. This means that they weren't up to the knowledge level expected before someone starts to learn about VoIP specifically.



  • @Dashrender said in Time to gut the network - thoughts?:

    I put 10 Gb on the table mainly because it's actual cost over 1 Gb links was around $100 more per link and it would give me a chance to - just do it.

    But as I mentioned above, when I put my IT hat one, you know, the one where we do what's best for the company - which includes being financially responsible - well, I suppose I should just save that $200 because it's really not going to gain us anything.

    It if REALLY gains you nothing, then yeah, that's just $200 wasted. Not a tonne, but not nothing either. Think if there is ever a time that having $200 extra in the budget would have been of value. Then think if this decision might make it harder to get needed money in the future - whether for IT or a new office chair or a desk or whatever.



  • @travisdh1 said in Time to gut the network - thoughts?:

    Are the fiber links setup with Link Aggregation? If not, that could make a big difference.

    No, but it really won't make that much difference from a throughput to the server, because the servers are limited to 1 Gb links too.

    The more we post, the more the 10 Gb links will be a complete waste unless I also install new NICs into my servers.



  • @Dashrender

    The more we post, the more the 10 Gb links will be a complete waste unless I also install new NICs into my servers.

    It's really nice when you have it, but costs a good bit.



  • @scottalanmiller said in Time to gut the network - thoughts?:

    @Dashrender said in Time to gut the network - thoughts?:

    Legacy understanding, and the belief (by the phone installation company) that VLANs would allow QOS for the IP phones.

    Should have walked them out the door the moment that you found out that they didn't know even the basic underpinnings of networking or phones. What value did they bring if they aren't aware of how either work? There are only two skills sets for VoIP to have... and thinking that VLANs do QoS indicates basically zero knowledge of either. That's super basic stuff. It doesn't require any special phone knowledge to know why that's impossible. This means that they weren't up to the knowledge level expected before someone starts to learn about VoIP specifically.

    Time out - while VLANs themselves don't do QoS, giving a VLAN a higher QoS over all other VLANs (which was their recommendation) does, right?



  • @scottalanmiller said in Time to gut the network - thoughts?:

    @Dashrender said in Time to gut the network - thoughts?:

    I put 10 Gb on the table mainly because it's actual cost over 1 Gb links was around $100 more per link and it would give me a chance to - just do it.

    But as I mentioned above, when I put my IT hat one, you know, the one where we do what's best for the company - which includes being financially responsible - well, I suppose I should just save that $200 because it's really not going to gain us anything.

    It if REALLY gains you nothing, then yeah, that's just $200 wasted. Not a tonne, but not nothing either. Think if there is ever a time that having $200 extra in the budget would have been of value. Then think if this decision might make it harder to get needed money in the future - whether for IT or a new office chair or a desk or whatever.

    Great points - and no, $200 spent now will have zero or near zero impact on my ability to spend in the future. Though, as additional posts point out, to make it have any real value, I would need to upgrade my VM hosts with 10 Gb NICs, so now the price is even higher, and while there would be a potential for a small amount of gain, the gain to value is really really low.



  • @Dashrender said in Time to gut the network - thoughts?:

    @scottalanmiller said in Time to gut the network - thoughts?:

    @Dashrender said in Time to gut the network - thoughts?:

    I put 10 Gb on the table mainly because it's actual cost over 1 Gb links was around $100 more per link and it would give me a chance to - just do it.

    But as I mentioned above, when I put my IT hat one, you know, the one where we do what's best for the company - which includes being financially responsible - well, I suppose I should just save that $200 because it's really not going to gain us anything.

    It if REALLY gains you nothing, then yeah, that's just $200 wasted. Not a tonne, but not nothing either. Think if there is ever a time that having $200 extra in the budget would have been of value. Then think if this decision might make it harder to get needed money in the future - whether for IT or a new office chair or a desk or whatever.

    Great points - and no, $200 spent now will have zero or near zero impact on my ability to spend in the future. Though, as additional posts point out, to make it have any real value, I would need to upgrade my VM hosts with 10 Gb NICs, so now the price is even higher, and while there would be a potential for a small amount of gain, the gain to value is really really low.

    Well in theory if you have two hosts and they were both "balls to the wall" with old 1Gb/s links and connected fully over the link you'd push into where there is "some" value. And presumably every host you have has multiple 1Gb/s links, so you need more than 1Gb/s "WAN" here to handle that.



  • @Dashrender said in Time to gut the network - thoughts?:

    @scottalanmiller said in Time to gut the network - thoughts?:

    @Dashrender said in Time to gut the network - thoughts?:

    Legacy understanding, and the belief (by the phone installation company) that VLANs would allow QOS for the IP phones.

    Should have walked them out the door the moment that you found out that they didn't know even the basic underpinnings of networking or phones. What value did they bring if they aren't aware of how either work? There are only two skills sets for VoIP to have... and thinking that VLANs do QoS indicates basically zero knowledge of either. That's super basic stuff. It doesn't require any special phone knowledge to know why that's impossible. This means that they weren't up to the knowledge level expected before someone starts to learn about VoIP specifically.

    Time out - while VLANs themselves don't do QoS, giving a VLAN a higher QoS over all other VLANs (which was their recommendation) does, right?

    Sure, QoS works regardless of the VLAN. So putting in a VLAN in order to get QoS is completely false.

    So there are two possibilities that I see...

    1. They aren't qualified to do what they are doing, even at the most basic "has never seen VoIP" level.
    2. Worse case, they DO know what they are doing and were actively making your environment more expensive and complex in order to charge you for more work (theft.)