The Cisco Short Questions Thread v12.4.9(WTF)

The Something Awful Forums > Discussion > Serious Hardware/Software Crap > The Cisco Short Questions Thread v12.4.9(WTF)

«‹›431 »

H.R. Paperstacks: May 1, 2006; This is America
My president is black
and my Lambo is blue

So what is everyone using for netflow collection and analysis? I'm in the process of talking to vendors because I am tired of dealing with cacti and it's inability to accurately get sub-5min spikes. Even with the adjustments for 1min polling, it's just a hack job. We mainly use cacti for historical purposes and capacity planning, but with the amount of data deal with, cacti has burned us a few times. The added features of netflow would allow better planning, but also analysis of what is actually going on.

So far I've looked at Scrutinizer(Plixer) and Solarwinds, and have an upcoming meeting with Netscout. It just seems like everyone only offers a Windows install, a VMware image, or an actual appliance. I'd really like something that I just sit on top of our already deployed RHEL6 boxes we use for other stuff on the management network.

Ntop is nice, but doesn't really provide historical ability since it is just real-time in memory and lost when rebooted. Unless that has changed since last time I messed with it.

# ? Mar 1, 2013 18:49

Adbot: ADBOT LOVES YOU

# ? May 14, 2024 16:15

psydude: Apr 1, 2008

WhatsUp Gold FlowMonitor is pretty damned good, but it doesn't run on RHEL.

# ? Mar 1, 2013 19:08

ate shit on live tv: Feb 15, 2004; by Azathoth

Does ior or someone have a custom captive portal page for a Wireless Lan Controller?

The one I have isn't working and I'm not a web dev, so I'm not sure what the problem is.

The Portal hsould just present the T&Cs then have an accept button. Currently this does all that, however the Accept button doesn't work, or at least it doesn't do anything.

Here is mine:

code:

<html>
<head>
<meta http-equiv="Pragma" content="no-cache">
<meta HTTP-EQUIV="Content-Type" CONTENT="text/html; charset=iso-8859-1">
<title>Welcome </title>
<script>

function submitAction(){
      var link = document.location.href;
      var searchString = "redirect=";
      var equalIndex = link.indexOf(searchString);
      var redirectUrl = "";

      if (document.forms[0].action == "") {
      var url = window.location.href;
      var args = new Object();
      var query = location.search.substring(1);
      var pairs = query.split("&");
          for(var i=0;i<pairs.length;i++){
              var pos = pairs[i].indexOf('=');
              if(pos == -1) continue;
				var argname = pairs[i].substring(0,pos);
				var value = pairs[i].substring(pos+1);
				args[argname] = unescape(value);
          }
          document.forms[0].action = args.switch_url;       
      }       
      if(equalIndex >= 0) {
            equalIndex += searchString.length;
            redirectUrl = "";
            redirectUrl += link.substring(equalIndex);
      }
      if(redirectUrl.length > 255)
      redirectUrl = redirectUrl.substring(0,255);
      document.forms[0].redirect_url.value = redirectUrl;
      document.forms[0].buttonClicked.value = 4;
      document.forms[0].submit();
}

function reject()
{
	alert("You will not be able to access the system!");
}


function loadAction(){
      var url = window.location.href;
      var args = new Object();
      var query = location.search.substring(1);
      var pairs = query.split("&");
      for(var i=0;i<pairs.length;i++){
          var pos = pairs[i].indexOf('=');
          if(pos == -1) continue;
          var argname = pairs[i].substring(0,pos);
          var value = pairs[i].substring(pos+1);
          args[argname] = unescape(value);
      }
      document.forms[0].action = args.switch_url;

}

</script>
</head>

<body bgcolor=#FFEECC topmargin="50" marginheight="50" onload="loadAction();"> <form method="post"> <input TYPE="hidden" NAME="buttonClicked" SIZE="16" MAXLENGTH="15" value="0"> <input TYPE="hidden" NAME="redirect_url" SIZE="255" MAXLENGTH="255" VALUE=""> <input TYPE="hidden" NAME="err_flag" SIZE="16" MAXLENGTH="15" value="0">

<table border="0" cellspacing="0" cellpadding="0">

<h1 align=center><font color="#336699">Welcome</font></h1>
<iframe src="./aup.html" width="800" height="500" scrolling="auto"></iframe>
<p><input type="button" name="Submit" value="Accept" class="button" onclick="submitAction();"></p>
<p><input type="button" name="Reject" value="Reject" class="button" onclick="reject();"></p>
</td>
</tr>
</table>
</div>

</form>
</body>
</html>

# ? Mar 1, 2013 20:02

Bluecobra: Sep 11, 2001; The Future's So Bright I Gotta Wear Shades

routenull0 posted:

So what is everyone using for netflow collection and analysis? I'm in the process of talking to vendors because I am tired of dealing with cacti and it's inability to accurately get sub-5min spikes. Even with the adjustments for 1min polling, it's just a hack job. We mainly use cacti for historical purposes and capacity planning, but with the amount of data deal with, cacti has burned us a few times. The added features of netflow would allow better planning, but also analysis of what is actually going on.

So far I've looked at Scrutinizer(Plixer) and Solarwinds, and have an upcoming meeting with Netscout. It just seems like everyone only offers a Windows install, a VMware image, or an actual appliance. I'd really like something that I just sit on top of our already deployed RHEL6 boxes we use for other stuff on the management network.

Ntop is nice, but doesn't really provide historical ability since it is just real-time in memory and lost when rebooted. Unless that has changed since last time I messed with it.

Someone here mentioned Observium which looks nice but I haven't had a chance to check it out yet. If you want the most accurate bandwidth monitoring, you will need to buy optical/copper taps which will split the signal to a monitoring device. This can get super expensive, but paying the money for a packet capture monitoring system has saved our rear end with troubleshooting issues countless times. A good example of this is detecting microburst traffic. With our appliance, I can see traffic bursting to over 200Mbs in 5ms intervals, but over the course of 1 second it typically hovers around 25-50Mbs. If I only could see 1 second then I would think that all I would need is a 100Mbs circuit instead of a 1Gb one. This means packets will get dropped and/or delayed which can be bad if you have traffic that is latency sensitive.

# ? Mar 1, 2013 20:25

H.R. Paperstacks: May 1, 2006; This is America
My president is black
and my Lambo is blue

Bluecobra posted:

Someone here mentioned Observium which looks nice but I haven't had a chance to check it out yet. If you want the most accurate bandwidth monitoring, you will need to buy optical/copper taps which will split the signal to a monitoring device. This can get super expensive, but paying the money for a packet capture monitoring system has saved our rear end with troubleshooting issues countless times. A good example of this is detecting microburst traffic. With our appliance, I can see traffic bursting to over 200Mbs in 5ms intervals, but over the course of 1 second it typically hovers around 25-50Mbs. If I only could see 1 second then I would think that all I would need is a 100Mbs circuit instead of a 1Gb one. This means packets will get dropped and/or delayed which can be bad if you have traffic that is latency sensitive.

That is exactly why we are moving away from cacti and to something more granular. I didn't think about splitting it off, which we already do for several other monitoring tools, mainly IDS/IPS. What appliance are you using if you don't mind me asking?

# ? Mar 1, 2013 20:29

DONT THREAD ON ME: Oct 1, 2002; by Nyc_Tattoo; Floss Finder

We contract a cisco guy, and I want to make sure I ask him to do the right thing. I'm currently setting up a wowza server/content management system at my work. Despite our (relatively low) bandwidth, and my protests, we are running the server internally on a connection that gets a lot of use.

Now, our wowza requirements are not going to be that intense, so this isn't the end of the world. It wont be streaming to hundreds of people at once, but rather, a few very important people. Even so, I want to make sure that our Cisco guy sets up some sort of QoS system that gives bandwidth priority to the wowza server.

I don't really know what enterprise/industrial cisco routers offer in this realm. Ideally, I'd like it to give priority to all traffic routed to the internal IPs/ports used by our server. Is this a straightforward thing to ask for? What terminology does cisco use, is it just QoS?

# ? Mar 1, 2013 20:33

GOOCHY: Sep 17, 2003; In an interstellar burst I'm back to save the universe!

routenull0 posted:

So what is everyone using for netflow collection and analysis? I'm in the process of talking to vendors because I am tired of dealing with cacti and it's inability to accurately get sub-5min spikes. Even with the adjustments for 1min polling, it's just a hack job. We mainly use cacti for historical purposes and capacity planning, but with the amount of data deal with, cacti has burned us a few times. The added features of netflow would allow better planning, but also analysis of what is actually going on.

So far I've looked at Scrutinizer(Plixer) and Solarwinds, and have an upcoming meeting with Netscout. It just seems like everyone only offers a Windows install, a VMware image, or an actual appliance. I'd really like something that I just sit on top of our already deployed RHEL6 boxes we use for other stuff on the management network.

Ntop is nice, but doesn't really provide historical ability since it is just real-time in memory and lost when rebooted. Unless that has changed since last time I messed with it.

We use Scrutinizer.

# ? Mar 1, 2013 20:51

falz: Jan 29, 2005; 01100110 01100001 01101100 01111010

routenull0 posted:

netflow collection and analysis

NfSen - http://nfsen.sf.net

# ? Mar 1, 2013 20:56

H.R. Paperstacks: May 1, 2006; This is America
My president is black
and my Lambo is blue

falz posted:

NfSen - http://nfsen.sf.net

If you have this deployed, what is your storage like given how many flows/per second you are collecting?

GOOCHY posted:

We use Scrutinizer.

So far Scrutinizer is in the lead, even if I have to deploy a virtual image, it pretty much does everything we need.

# ? Mar 1, 2013 21:05

ragzilla: Sep 9, 2005; don't ask me, i only work here

routenull0 posted:

If you have this deployed, what is your storage like given how many flows/per second you are collecting?

So far Scrutinizer is in the lead, even if I have to deploy a virtual image, it pretty much does everything we need.

peak 36k flow/sec (average more like 20-25k flow/sec) stores 5-6 days of flows in 250GB. Netflow won't really detect microbursts that well- SNMP based graphing is still one of the better ways to detect it- but Netflow will let you determine what caused it.

Have you checked what the minimum counter update time on your platform is?

# ? Mar 1, 2013 21:32

Yeast Confection: Oct 7, 2005

routenull0 posted:

So what is everyone using for netflow collection and analysis?

We use Solarwinds, but its primary function for us is SNMP polling/alerting, config management and (soon) IP management. Our using it for netflow was kind of an afterthought. I find it easy to use. They've been making a lot of improvements to their products over the past few years.

# ? Mar 1, 2013 21:38

jwh: Jun 12, 2002

chumpchous posted:

I don't really know what enterprise/industrial cisco routers offer in this realm. Ideally, I'd like it to give priority to all traffic routed to the internal IPs/ports used by our server. Is this a straightforward thing to ask for? What terminology does cisco use, is it just QoS?

It is straightforward, but be mindful that you have very limited ability to implement QoS in the inbound direction. Once you've received a packet via your NSP, the only forwarding decisions you can make are to forward the packet or to drop it (for the sake of simplicity).

Where you'll want a more robust bandwidth guarantee is in your outbound direction, from your server to the NSP. This is both reasonable and fairly easy to accomplish.

Remember the golden rule of QoS: QoS doesn't make something better, it makes other things worse.

If your traffic to the server is largely data asymmetric, as most webservers are, then I think a QoS policy in the outbound direction is reasonable.

If that's not the case, and your users will be sending the majority of traffic to the server, you're going to be exceptionally limited as to what you can do.

# ? Mar 1, 2013 22:32

Bluecobra: Sep 11, 2001; The Future's So Bright I Gotta Wear Shades

routenull0 posted:

That is exactly why we are moving away from cacti and to something more granular. I didn't think about splitting it off, which we already do for several other monitoring tools, mainly IDS/IPS. What appliance are you using if you don't mind me asking?

We use a combination of Gigamon optical/copper taps to "split" our WAN links into their traffic aggregator (GigaVUE-420) and then we feed this into two 10Gb ports into our Corvil appliance. Corvil is targeted mainly towards financial companies and it carries a huge price tag. At end of the day, you will likely be spending over $100K for all the Gigamon stuff and a 10Gb Corvil appliance. Corvil is a spin-off company that used to be Cisco Bandwith Quality Monitor (BQM).

# ? Mar 1, 2013 22:34

jwh: Jun 12, 2002

Re: NetFlow

Nothing is very good, but it's better than nothing. Plus, the port/application mentality is only of limited usefulness these days. I used to hate seeing tcp/80 as a top bandwidth consumer, because that can be just about anything from an application perspective.

I hate to keep sounding like a salesman for Palo Alto, but their port agnostic App-ID is a great deal better than NetFlow for correlation purposes.

edit: and I was surprised recently to discover a PA-500 can be had with all the optional feature licensing (IPS/URL/Wildfire) for about $5k. I had never quoted the low end of their product before, and this came as a nice surprise.

# ? Mar 1, 2013 22:35

Mierdaan: Sep 14, 2004; Pillbug

jwh posted:

edit: and I was surprised recently to discover a PA-500 can be had with all the optional feature licensing (IPS/URL/Wildfire) for about $5k. I had never quoted the low end of their product before, and this came as a nice surprise.

I have a real feeling you guys are going to sell me on something like this to replace our ASA 5505. Those things are cheap as hell, but the downsides are numerous and just talking about netflow wants me make to hit it with a hammer.

# ? Mar 1, 2013 22:48

DONT THREAD ON ME: Oct 1, 2002; by Nyc_Tattoo; Floss Finder

jwh posted:

If your traffic to the server is largely data asymmetric, as most webservers are, then I think a QoS policy in the outbound direction is reasonable.

If that's not the case, and your users will be sending the majority of traffic to the server, you're going to be exceptionally limited as to what you can do.

Great. We do have a lot of inbound traffic, but I'm not interested in restricting it; outbound is my major concern. I just wanted to be sure I wasn't asking for the impossible.

Can the traffic policy be shaped dynamically? For instance, can I allow people more or less unrestricted outgoing bandwidth, until there is demand from the wowza server, or do I have to define static limits?

DONT THREAD ON ME fucked around with this message at 23:10 on Mar 1, 2013

# ? Mar 1, 2013 23:07

jwh: Jun 12, 2002

Mierdaan posted:

I have a real feeling you guys are going to sell me on something like this to replace our ASA 5505. Those things are cheap as hell, but the downsides are numerous and just talking about netflow wants me make to hit it with a hammer.

Well, even with additional user licensing (which is stupid, imo), an ASA 5505 is going to cost 1/10th of a PA500, for example. You may want to look even lower in the product spectrum, at the PA200. They have a similar form-factor to the ASA 5505, also, if that's important.

I can tell you this, for what it's worth: PA is very much money well spent if you're looking for additional visibility and control over user networks.

# ? Mar 1, 2013 23:25

jwh: Jun 12, 2002

chumpchous posted:

Great. We do have a lot of inbound traffic, but I'm not interested in restricting it; outbound is my major concern. I just wanted to be sure I wasn't asking for the impossible.

Can the traffic policy be shaped dynamically? For instance, can I allow people more or less unrestricted outgoing bandwidth, until there is demand from the wowza server, or do I have to define static limits?

It can be shaped dynamically, yes. Though typically the way this would work would be to assign from the server to a queue with a greater bandwidth reservation than the rest of your commodity data. In cases where your commodity data is the only thing on the wire, it will have access to the full bandwidth, and when contention occurs, the bandwidth reservation will favor the server (to a configured point).

Like all things Cisco and particularly Cisco QoS, there's about a dozen different ways to achieve nearly the same result, so you may just want to defer to your contractor on the specific implementation.

If your Cisco contractor says it can't be done have him look up something like 'Hierarchical Queuing Framework' / MQC.

# ? Mar 1, 2013 23:29

H.R. Paperstacks: May 1, 2006; This is America
My president is black
and my Lambo is blue

The added analysis (top talkers, ports/protocol) after collecting the netflow are just a bonus for us. My real concern is historically tracking bandwidth usage and cacti has been fine, but there have been several instances where cacti will be reporting 80% utilization of a link, but in fact, due to the nature of our data flows and collection, we regularly peak the circuit. It just doesn't show up because of polling intervals or the device only having 5min counters.

# ? Mar 1, 2013 23:38

jwh: Jun 12, 2002

In that case I would expect NetFlow to be even less useful. By default, it's what, 15 minute exports? 5? I forget.

Every NetFlow graph I've ever seen has looked like a row of triangular circus tents.

# ? Mar 2, 2013 00:04

GOOCHY: Sep 17, 2003; In an interstellar burst I'm back to save the universe!

I think it's 5 minute average by default. We just use it to capture source/dest traffic data and push it back to the client saying, "Stop complaining that it's slow - you have 10 people streaming Pandora and they're pegging the bandwidth out."

# ? Mar 2, 2013 00:09

DONT THREAD ON ME: Oct 1, 2002; by Nyc_Tattoo; Floss Finder

jwh posted:

It can be shaped dynamically, yes. Though typically the way this would work would be to assign from the server to a queue with a greater bandwidth reservation than the rest of your commodity data. In cases where your commodity data is the only thing on the wire, it will have access to the full bandwidth, and when contention occurs, the bandwidth reservation will favor the server (to a configured point).

Like all things Cisco and particularly Cisco QoS, there's about a dozen different ways to achieve nearly the same result, so you may just want to defer to your contractor on the specific implementation.

If your Cisco contractor says it can't be done have him look up something like 'Hierarchical Queuing Framework' / MQC.

Thanks, I'm basically just forwarding this to our cisco guy!

# ? Mar 2, 2013 00:48

bort: Mar 13, 2003

How prevalent is the problem of microbursting? I've heard it blamed for a few errors in my infrastructure, but that always turned out to be something else. However, with the growth of 10G networking I could see situations where it'd occur. Is it a characteristic of app traffic or just volume?

We use Riverbed Cascade for netflow, but I don't think it'd make sense as a solution without a prior investment in Steelheads. Leveraging them for traffic analysis is rewarding, e: but we're in the same boat as everyone else when the traffic doesn't cross one.

Bluecobra, I'm jealous of your solution.

bort fucked around with this message at 02:27 on Mar 2, 2013

# ? Mar 2, 2013 02:24

ragzilla: Sep 9, 2005; don't ask me, i only work here

bort posted:

How prevalent is the problem of microbursting? I've heard it blamed for a few errors in my infrastructure, but that always turned out to be something else. However, with the growth of 10G networking I could see situations where it'd occur. Is it a characteristic of app traffic or just volume?

If you have many-to-one traffic patterns (aggregation/access devices) or going from higher speed to lower speed (distribution devices- even more fun with cut-through!) microbursts can irritate you. On the other hand the packets dropped due to the burst do get the app to slow down if it's well behaved.

# ? Mar 2, 2013 03:08

teh z0rg: Nov 17, 2012

Bluecobra posted:

If you want the most accurate bandwidth monitoring, you will need to buy optical/copper taps which will split the signal to a monitoring device.

Do not do this - do not tap optical and steal your light. Port mirror that poo poo on good hardware like a Juniper Networks device.

teh z0rg fucked around with this message at 04:30 on Mar 2, 2013

# ? Mar 2, 2013 04:26

teh z0rg: Nov 17, 2012

routenull0 posted:

So what is everyone using for netflow collection and analysis?

Arbor Peakflow.

# ? Mar 2, 2013 04:26

doomisland: Oct 5, 2004

teh z0rg posted:

Arbor Peakflow.

Wicked jealous.

# ? Mar 2, 2013 04:30

ragzilla: Sep 9, 2005; don't ask me, i only work here

teh z0rg posted:

Arbor Peakflow.

Doing TMS as well, or just flow analysis?

teh z0rg posted:

Do not do this - do not tap optical and steal your light. Port mirror that poo poo on good hardware like a Juniper Networks device.

Unless you have a power meter, and understand your link budgets, and you're only contemplating doing it on SM fiber. Only costs ~3dBm to do a 2-way split, you'd still have at least 6.5dBm of budget on LX with a 2-way.

# ? Mar 2, 2013 05:51

Bluecobra: Sep 11, 2001; The Future's So Bright I Gotta Wear Shades

bort posted:

How prevalent is the problem of microbursting? I've heard it blamed for a few errors in my infrastructure, but that always turned out to be something else. However, with the growth of 10G networking I could see situations where it'd occur. Is it a characteristic of app traffic or just volume?

We use Riverbed Cascade for netflow, but I don't think it'd make sense as a solution without a prior investment in Steelheads. Leveraging them for traffic analysis is rewarding, e: but we're in the same boat as everyone else when the traffic doesn't cross one.

Bluecobra, I'm jealous of your solution.

In the financial industry it is a big problem. Market data tends to be a shitton of small bursty packets. If your network can't handle these short traffic bursts, that could mean a delayed order or a price update and that is bad news for trading. Here is some good info about microbursts:

http://www.latencystats.com/blog/market-data-microbursts
http://www.latencystats.com/blog/provisioning-microbursts
http://blog.endace.com/2012/01/what-is-a-microburst-really/

teh z0rg posted:

Do not do this - do not tap optical and steal your light. Port mirror that poo poo on good hardware like a Juniper Networks device.

The problem with port mirroring/SPAN is the added latency which skews your results. Also, on most switches I have seen there are pretty low limits to how many sessions you can have. A Cisco 6500-E with a Sup720 can only support 2 span sessions AFAIK. The Cisco Nexus 3548 looks to be one of the better options to span traffic (150ns latency), but again there are so many sessions you can have. If you're in a co-location facility and have 20+ cross-connects and you want to capture that data, taps/aggregation devices are your only way.

Bluecobra fucked around with this message at 17:05 on Mar 2, 2013

# ? Mar 2, 2013 16:59

teh z0rg: Nov 17, 2012

Bluecobra posted:

The problem with port mirroring/SPAN is the added latency which skews your results. Also, on most switches I have seen there are pretty low limits to how many sessions you can have. A Cisco 6500-E with a Sup720 can only support 2 span sessions AFAIK. The Cisco Nexus 3548 looks to be one of the better options to span traffic (150ns latency), but again there are so many sessions you can have. If you're in a co-location facility and have 20+ cross-connects and you want to capture that data, taps/aggregation devices are your only way.

Which is why i said Juniper. :smug:

# ? Mar 2, 2013 17:12

Thanks Ants: May 21, 2004; #essereFerrari

lol internet. posted:

I know this is a Cisco thread but I didn't want to create a new thread for this question and since Cisco bought out Meraki, I figured I'd ask here.

I'm probably going to get a Meraki MX60 router. I was wondering if anyone has experience with any of their products. I'm wondering if I can just use it as a home router with the features, or will I need the subscription licenses? I've been having trouble finding info on this online.

You need the license for it to be usable. I've had some Meraki kit, the MX60 arrived with a dead port and it took a while to convince the support guy I wasn't an idiot and to swap it. Then an MR16 died and it took two weeks to get the replacement shipped.

The management is nice and it's really easy, but the quality control of the products and the support sucks poo poo.

# ? Mar 2, 2013 17:55

Sepist: Dec 26, 2005; FUCK BITCHES, ROUTE PACKETS; Gravy Boat 2k

Bluecobra posted:

The problem with port mirroring/SPAN is the added latency which skews your results. Also, on most switches I have seen there are pretty low limits to how many sessions you can have. A Cisco 6500-E with a Sup720 can only support 2 span sessions AFAIK. The Cisco Nexus 3548 looks to be one of the better options to span traffic (150ns latency), but again there are so many sessions you can have. If you're in a co-location facility and have 20+ cross-connects and you want to capture that data, taps/aggregation devices are your only way.

We use VACL's with the capture command to mirror traffic to our infinistream console, mostly the same thing with minor differences.

# ? Mar 2, 2013 18:01

Bluecobra: Sep 11, 2001; The Future's So Bright I Gotta Wear Shades

teh z0rg posted:

Which is why i said Juniper.

I just looked at the QFX3500 and it said it only supports 4 mirroring sessions. Do they make any non-chassis 10GbE switches that support more sessions? Apart from our MAG Gateways (which rock), I am not too familiar with what stuff Juniper has.

# ? Mar 2, 2013 18:05

teh z0rg: Nov 17, 2012

EX4500 have 7 port based and one global

I assume the new EX4550 has the same or more.

# ? Mar 2, 2013 18:57

ate shit on live tv: Feb 15, 2004; by Azathoth

teh z0rg posted:

Do not do this - do not tap optical and steal your light. Port mirror that poo poo on good hardware like a Juniper Networks device.

Passive optical splitting is fine. Short Ranges, matched optics, you'll never have a problem. It can be MM or SM really doesn't matter, it's physics and a "stronger" light level doesn't mean more bandwidth.

# ? Mar 2, 2013 21:27

optik: Jul 6, 2005; linux is a pathway to many abilities..... some consider to be un-natural

To jump on the bandwagon another reason that proper taps/splitters are a better alternative is that for monitoring issues (not so much microspikes) most platforms won't mirror/SPAN frames that are considered invalid and will be dropped, if you were just relying on SPAN sessions to see this then you would be SOL

# ? Mar 3, 2013 02:00

teh z0rg: Nov 17, 2012

Powercrazy posted:

Passive optical splitting is fine. Short Ranges, matched optics, you'll never have a problem. It can be MM or SM really doesn't matter, it's physics and a "stronger" light level doesn't mean more bandwidth.

more power = better than

# ? Mar 3, 2013 03:01

adorai: Nov 2, 2002; 10/27/04 Never forget; Grimey Drawer

teh z0rg posted:

more power = better than

There is a power level in which it works. Anything above that level is good enough.

# ? Mar 3, 2013 04:55

BelDin: Jan 29, 2001

Powercrazy posted:

Does ior or someone have a custom captive portal page for a Wireless Lan Controller?

The one I have isn't working and I'm not a web dev, so I'm not sure what the problem is.

The Portal should just present the T&Cs then have an accept button. Currently this does all that, however the Accept button doesn't work, or at least it doesn't do anything.

I think I've still got mine around if nobody posts one by tomorrow. Out of town at the moment.

# ? Mar 3, 2013 20:07

Adbot: ADBOT LOVES YOU

# ? May 14, 2024 16:15

ate shit on live tv: Feb 15, 2004; by Azathoth

BelDin posted:

I think I've still got mine around if nobody posts one by tomorrow. Out of town at the moment.

I actually figured it out. It had nothing to do with the webpage, I need to change the Web authentication to "passthrough" and that fixed it.

# ? Mar 4, 2013 02:04

The Something Awful Forums > Discussion > Serious Hardware/Software Crap > The Cisco Short Questions Thread v12.4.9(WTF)

«‹›431 »