 TomS_Git-r-done
MVM
join:2002-07-19
London, UK |
to DarkLogix
Re: [Config] Performance due to VLAN, incomplete STP, or loop backs?He has 2960S switches, not 2960G talked about in that thread. |
|
 DarkLogixTexan and Proud
Premium Member
join:2008-10-23
Baytown, TX |
DarkLogix
Premium Member
2013-Dec-10 10:28 am
said by TomS_:He has 2960S switches, not 2960G talked about in that thread. And the cisco page I skimmed yesterday included the G and the S in supporting that feature But I forgot the IOS ver that it mentioned it being added in. why do you assume badly of me? |
|
| |
to HELLFIRE
2960S can stack 4 and we deploy them with total of 5 VLANs. stale,user, printer,mgmt,voice. 2960S is utter junk 2975s on the other hand are awesome other than it cannot support POE on all ports. But the QoS on the stack and other factors are there... |
|
| Da Geek Kid |
to DarkLogix
said by DarkLogix:Oh and we've had the EX4200's still have the same happen after going to 11.
the guy that normally messes with them was out so I had to reload the OS via serial. We have over 50 4200s (non)stacked and many different configs including MPLS. We have had 0 down time on them. Now that you brought it up, we need to update the code since it's been a long while we touched them. they have been up for over a year. |
|
|
DarkLogixTexan and Proud
Premium Member
join:2008-10-23
Baytown, TX
2 edits |
I wish my experience with them were even remotely close.
we now have a stack of 3x EX4200-48P with 2x EX4200PX (any idea whats new with the X?) though the price dropped greatly from when the 3 were bought |
|
| |
to Da Geek Kid
The 2975 has gone EOL and is replaced by the 2960S, which is soon to be replaced by the 2960X.
As mentioned earlier, just how much were you asking the 2960 to do before determining it utter junk? I have had 0 reliability issues with mine in the field. For reference, they are all doing inter vlan routing for the LAN's for between 5 - 10 VLAN's and a little QoS/VoIP. A couple of them have ACL's / VLAN maps for inter VLAN security but that's it. No dot1x, no mac address security, no DHCP Snooping. Most have to push high end streamed media, I'm talking 1080p 3D and soon, 4K content. They seem to handle it just fine. |
|
DarkLogixTexan and Proud
Premium Member
join:2008-10-23
Baytown, TX |
Finding out that the 2960 can do static routing has me interested, but I'm not sure how I'd want to do it
with a 3560 I'd be able to do it with a routing protocol, but with a 2960 I'd have to use static routes, so I goues I'd have to make a static rout pointing to each of the 4 etherswitch's p-t-p links and have them all with equal metrics but would those 4 be last resort routes or would I then need more routes to cover that.
(I'm not a fan of static routes because you have to make so many of them to have route redundancy, and well will a route load balance static routes?)(although I'd have to get my hands on that newer IOS that adds this support.) |
|
| |
said by DarkLogix:I'm not a fan of static routes because you have to make so many of them Summarise... although to be fair you still have to add an SVI for each VLAN so there are no shortcuts here. But once the SVI and layer 2 VLAN is up it'll appear in the routing table so there is no "extra" work to do unless you force it upon yourself. No idea how you would interact with the etherswitch on the 3845 'cos I've never used one. I doubt it would take long to figure it out in the lab though. It's simple routing after all. In production, as I have said before, I use the 2960S to route local LAN / inter-VLAN traffic with a default route pointing to the edge router. This keeps load off the router so it can concentrate on the various security and NAT duties required for internet connectivity. I have found it works very well indeed. |
|
DarkLogixTexan and Proud
Premium Member
join:2008-10-23
Baytown, TX |
I have 3 (planned 4) etherswitches each is running the same IOS as a 3750 with AdvipServ IOS, and are able to do intervlan routing, what I'd like is to load balance them at L3 so I'd need 4 static routes and for them to balance the load |
|
| |
to tubbynet
"so i must be smarter than anyone in the room" I clean up a lot after "consultants" and "IT" guys because they were a little smarter than everyone else in the room.  |
|
| battleop |
to aryoba
That's the problem with dealing with "Consultants" and not real engineers. |
|
| battleop |
to tubbynet
When you hear terms synergy like that RUN!!!!!!!!! |
|
| battleop |
to tubbynet
Just refrain replacing yes with ABSOLUTELY!! and cut back on using actually and you may be OK. |
|
DarkLogixTexan and Proud
Premium Member
join:2008-10-23
Baytown, TX |
DarkLogix
Premium Member
2013-Dec-11 10:22 am
Thought about it more and well using a 2960 with static only routing would not be so good for my setup
1. I'd have to put in 4x 0.0.0.0 routes on it then I'd have to put in 8 static routes on each etherswitch (though maybe put them on one and do redistribute statics) and then static routes don't react to link state changes so it'd not be so good. |
|
 tubbynetreminds me of the danse russe
MVM
join:2008-01-16
Gilbert, AZ |
said by DarkLogix:then static routes don't react to link state changes so it'd not be so good. ipsla with object tracking is your friend here. doubt the c2960 support it, but you can inject reachability, etc into decision making for injection/withdrawl of routes, etc. q. |
|
DarkLogixTexan and Proud
Premium Member
join:2008-10-23
Baytown, TX |
DarkLogix
Premium Member
2013-Dec-11 10:42 am
I think I'm leaning more toward wanting a WS-C3560G-24PS so I can just do dynamic instead
(as I doubt I'll get that newer IOS anyway)
Though I guess I'll look into smartnet for my 2960G to see if its in a price range I might be able to do |
|
tubbynetreminds me of the danse russe
MVM
join:2008-01-16
Gilbert, AZ |
said by DarkLogix:I think I'm leaning more toward wanting a WS-C3560G-24PS so I can just do dynamic instead it may be easier to start building a lab so you don't muck up your production network. q. |
|
DarkLogixTexan and Proud
Premium Member
join:2008-10-23
Baytown, TX |
DarkLogix
Premium Member
2013-Dec-11 10:48 am
Its my home network, so it puts some pressure on me to actually solve any issue I come to where as in a lab I might just get lazy.
well looks like about $200 for smartnet for the 2960G (CON-SNT-C2960G2C) vs $800 ebay for the 3560G
reasons for the 3560G that I searched
gig POE (could upgrade my AP without having to go to an injector)
dynamic routing
less limited
reasons for the Smartnet
cheaper
upgrade the IOS on the 2960G to gain new features |
|
tubbynetreminds me of the danse russe
MVM
join:2008-01-16
Gilbert, AZ |
said by DarkLogix:Its my home network, so it puts some pressure on me to actually solve any issue I come to where as in a lab I might just get lazy. yes -- but a lab allows you to actually think through a problem and not only find a solution, but understand *why* it works, rather than running to forums and going "OH NOES!!! TEH INTERNETZ IS BROKE!!!11". just a thought. q. |
|
DarkLogixTexan and Proud
Premium Member
join:2008-10-23
Baytown, TX |
DarkLogix
Premium Member
2013-Dec-11 11:06 am
Well I only posted that NAT post because I looked and couldn't find anything
I did clearly have a workaround to get on the internet. |
|
| |
px is the poe+ on the EX switches... The 2960S has been as every can see has gone thru it's short life span that was well over due. I am not putting my money on the 2960X either. The Stack needs to have QoS, 2960 does not. Anyone saying they have no problems on a stack of 2960s, then they really are not using it. I call the SOHO users. It's designed for such, not enterprise class. |
|
| |
to Nsane_iceman
Not to put too fine a point on it, but "hijacking OP's original thread...." COMPLETED!  Regards |
|
| |
with a complete page worth.. LOL! |
|
DarkLogixTexan and Proud
Premium Member
join:2008-10-23
Baytown, TX |
to HELLFIRE
said by HELLFIRE:Not to put too fine a point on it, but "hijacking OP's original thread...."
COMPLETED!
Regards What yo talkin bought willis? this is grand theft thread. |
|
 Nsane_iceman
Premium Member
join:2001-02-26
North Richland Hills, TX |
to HELLFIRE
said by HELLFIRE:Not to put too fine a point on it, but "hijacking OP's original thread...."
COMPLETED!
Regards Yup, little bit  |
|
DarkLogixTexan and Proud
Premium Member
join:2008-10-23
Baytown, TX |
said by Nsane_iceman:said by HELLFIRE:Not to put too fine a point on it, but "hijacking OP's original thread...."
COMPLETED!
Regards Yup, little bit Well thats what happens when the OP doesn't post for awhile. |
|
Nsane_iceman
Premium Member
join:2001-02-26
North Richland Hills, TX |
to Da Geek Kid
said by Da Geek Kid:just to clarify something... Switch 1 and Switch 2 are Core switches... Switch 6 and partial switch 7 config show they are just layer 2 switches... 3560's have plenty of beef to get this tiny network going.
what is 10.16.255.6??? My "core" and the thread's core differ. "Core" to me is can't run without it, while here it seems to real definition to core is no access devices (printer, desktop, server, AP) attached. I really don't have a core switch. ip route 0.0.0.0 0.0.0.0 10.16.255.6
ip route 10.0.0.0 255.0.0.0 10.16.254.1
ip route 10.19.0.0 255.255.0.0 10.16.255.6
ip route 10.21.0.0 255.255.0.0 10.16.255.6
ip route 10.25.0.0 255.255.0.0 10.16.255.6
ip route 10.26.0.0 255.255.0.0 10.16.255.6
ip route 10.28.0.0 255.255.0.0 10.16.255.6
ip route 10.30.0.0 255.255.0.0 10.16.255.6
ip route 10.100.0.0 255.255.0.0 10.16.255.6
If I'm understanding it correctly, I think 10.16.255.6 is the inside interface of our MIS and 10.16.254.1 is the inside of the MPLS. ...but, either I'm wrong or it's configured incorrectly or most likely it's both... ip route 0.0.0.0 0.0.0.0 10.16.255.6 = anything "unknown" route to 10.16.255.6 ip route 10.0.0.0 255.0.0.0 10.16.254.1 = anything going to 10.*.*.* go to 10.16.254.1 to go out ip route 10.19.0.0 255.255.0.0 10.16.255.6 = another location out of state, wouldn't be going out MIS, must be the inside interface of the PIX 506E that I didn't think was hooked up because from what I could tell both interfaces were plugged into the same switch. |
|
| Nsane_iceman |
to HELLFIRE
said by HELLFIRE:@Nsane_iceman
Make sure any interfaces in half duplex aren't an uplink to another connecting switch... just to be sure.
And now that I look at Da Geek Kid  's question and your configs, what's up with all the static routes?
If Switch01 and Switch06 are your "core" switches with the VLAN SVIs, why are you pointing everything at that
address? Methinks you may want to redo your diagram WITH the interface-to-link status AND an addressing scheme
as well.
Regards All interfaces in half duplex are conveyor belt PLC units, no uplinks. |
|
| Nsane_iceman |
to Da Geek Kid
said by Da Geek Kid:switch 6 has one vlan and it's not considered a core, but rather L2 switch. Another Major issue that was mentioned is that the core switches 1 and 2 do not have the same IOS which will break the network if configured improperly... In this situation, since there are no routing protocols running on either switch, it should be ok. All I can say is, 100% pure amateur design "Foe-Sho!" I've emailed Netech to see if they can send me the advipservices image for all the switches. Hopefully they can and not $160/hr for them to do it on all that lack it. |
|
DarkLogixTexan and Proud
Premium Member
join:2008-10-23
Baytown, TX
2 edits |
to Nsane_iceman
By definition there are 3 levels of switch position and /or role
Core, Distribution, and Access
Core - central main switch(s)
Distribution - as the name says are to distribute network links out
Access - end devices connect
in a small network the 3 collapse into one and as a network grows the need for a separation grows
So say you had a really large network, then maybe you'd have a core switch with 100gb interfaces then that would link to some lesser switches that have 100gb but only as uplinks and have 10gb mostly, then you have access switches uplinked to the distribution via 10gb and providing 1gb interfaces for end devices
in such a network servers might link directly to the distribution switches (as 100gb interfaces are not likely to be found for servers yet) but servers could easily have some 10gb nic's
do your PLC's not support full duplex?
I recall in working with some AB SLC500's (up to 505's) that I was able to get them to do 100mb FD as well as the controllogix ethernet boards, or are you just leaving PLC stuff to the PLC people which might not know anything about Half vs full.
IIRC the 501-504's had to have a serial to ethernet device added but the 505's had it built in, made it easier as the 505's I could just plug in then open RS Logix 500 and set the IP and interface over the serial, where as on the 500-504's I had to setup the serial-eth device then set the PLC to know about it. |
|
