Podcast #128: We chat with Kent C Dodds about why he loves React and discuss what life was like in the dark days before Git. Listen now.

Hot answers tagged

19

This can introduce a number of problems, like additional attenuation or cross talk. Splicing is to be avoided whenever possible, but I have seen this work in a pinch although I would never recommend it. They key is to use a cable certification tester (not just a continuity tester) to make sure it still passes your required standard (Cat5/5E/6) after ...


15

You are totally correct, if we have to use an instruction cycle per bit then 10Gbps would be unachievable. So the first thing to note is that we handle a word per CPU instruction -- 64 bits. Even then the worst thing we can do for performance is to have the CPU access all of the words of a packet. Thus the focus on "zero-copy" handling of packets. Some of ...


13

Neither, really. Replacing a copper link with a fiber link might lower latency a tiny bit (assuming an uncongested link), but what you really get when you replace a "core" link with a higher bandwidth link is less possible congestion. In your example scenerio, it doesn't matter, because there is only one device on each end. In a live network, though, moving ...


11

An an example Juniper MX80 has ingress->egress delay of about 8us, on low-latency cut-through switch it can be <1us (maybe 0.7us). (Remember that cut-through switch can't do cut-through 100% of time, only when egress port happens to be idle!) 1km in fibre is about 5us latency (again, single direction). Serialization delay @10G for minimum sized payload (...


8

Short answer: No, ISPs do not use TCP Slow-Start as the main way to throttle customer bandwidth. Long answer: There are many ways that an ISP can throttle the bandwidth a particular customer gets allotted, but in my experience tweaking TCP knobs is very far down the list of possible methods. If you are an ISP, you can control the traffic volume at a lower ...


8

If you're trying to test 1xGE No Drop Rate and measure circuit delay within 8ms, I would use nuttcp to test bandwidth and iperf2 / mtr to test delay. I would do the following... Find two linux desktop PCs, if possible (laptops are sometimes acceptable, but you might run into issues with chipset or bus performance at 1GE speeds). You can boot into a ...


7

The command show interface should be showing you the information you are looking for. router#show ver | i 29 Cisco IOS Software, C2900 Software (C2900-UNIVERSALK9-M), Version 15.3(2)T, RELEASE SOFTWARE (fc3) router#show int | i Giga.*up|Duplex GigabitEthernet0/0 is up, line protocol is up Full Duplex, 100Mbps, media type is RJ45 GigabitEthernet1/0 is up, ...


7

The usual suspects in the case of autonegotiations are: bad cables bad ports, SFPs or NICs port configs (are they both set to auto and do they both list the 1000BASE-T FD ability?) bugs, compatibility or legacy issues I would not recommend a fixed speed and duplex setting in this (or almost any) situation. The days of frequent autoneg issues and ...


7

You are conflating many things here, so let's try to detangle the issues in your question. Data rate is data rate, regardless of the physical medium. A 1Gb connection has the same data rate whether it is fiber or copper. As @toddwilcox mentions, the advantages of fiber over copper are longer spans and electromagnetic isolation. Data rates are independent ...


7

That's by product design. ISR 4k come with a platform shaper (upgradeable to ~2x the value by license upgrade with the "PERF" license). Cisco say that the limits of the platform shapers can be fully exploited, no matter how many features you turn on: NAT, QoS, IPSec, WAN Acceleration etc. Performance/throuhgput estimations had always been a bit if a story ...


6

UDP by itself isn't reliable. The data acknowledge/retransmission functions have to occur at a higher (i.e. application) level. TFTP is a good example of that.


6

The speed of the flow of data makes no difference in the physics of the medium. By this I mean that it takes the same time for an electric signal to flow from one side of a 100 meter copper run to the other, no matter if that signal is part of a 10Mbps or a 1Gbps link. If you change from copper to fiber, then you may notice a small improvement, but it ...


6

Cat6 can provide Gigabit speeds up to 90+ meters. It is far more likely that one of the terminations is the problem. Gigabit requires the use of all 8 wires, so if one (or both) of the terminations isn't making good contact, then it could run at 100M which only requires 4 wires.


6

If you have PCs at both ends then you could run xjperf, Qcheck from Ixia or other tools. You might get different results depending on if you use UDP or TCP and the number of sessions. For a distance over 100 miles you are looking at a minimum RTT of 1.6 ms at the speed of light in fibre/copper. So your RTT should be very low, maybe only a couple of ms. Say ...


6

iperf can do that job. Just ensure you test your devices back to back first so you know the capabilities of the devices. Of course there are more professional tools as well. This is a good RFC to read up proper methology: http://www.ietf.org/rfc/rfc2544.txt


6

The -l option is for the buffer and doesn't influence the amount of data transferred. You have to specify the desired amount of data with the client-only option -n in KByte or MByte. So for 10GB, use -n 10240M Example: With the defaut buffer size of 8KB: iperf -c 10.1.1.1 -n 10240M ------------------------------------------------------------ Client ...


6

In reality, the legacy 10 Mbps ethernet interface probably can't negotiate, and it can probably only do half duplex (very few 10 Mbps interfaces can do full duplex). You should let the 1 Gbps interface auto-negotiate. It will try to negotiate, but if the 10 Mbps can't negotiate, it will detect (not negotiate) that the connection is 10 Mbps, and it will set ...


6

Fiber media converter is a small device with two media-dependent interfaces and a power supply, simply receive data signals from one media, convert and transmit them to another media. It can be installed almost anywhere in a network. The style of connector depends on the selection of media to be converted by the unit. The most common being UTP to multimode ...


5

In this case, moving from 1G end-to-end, to a 10G core shouldn't significantly change anything. Only a marginal increase in throughput would come from the faster signalling (decreased bit-time) on the 10G+ link. But in the absence of any congestion (read: other hosts), they should've been able to saturate the link to begin with. The time it takes hosts A &...


5

When you are using a wifi adapter in monitor mode, you are not associated to an access point. It will simply listen on the configured channel(s) and capture any traffic it understands whether that is 802.11b, 802.11g, or 802.11n. If you are associated to an access point, then you are running in promiscuous mode, and then you are generally only going to ...


5

The very short answer: don't configure anything. Auto negotiation (or the lack thereof as Ron's detailed) works only when it's left alone. Manual settings can very easily cause problems either right away when done incorrectly or later on when hardware is upgraded. For 1000BASE-T, auto negotiation is required - most hardware won't allow you to manually ...


5

The equipment interfaces determine the speed, albeit the medium used may have restrictions on which interfaces it can be used. If you have 1 Gbps ethernet, it doesn't matter if it is copper or fiber, it will be 1 Gbps ethernet. Properly installed and tested Category-6 cabling can even run 10 Gbps over a short distance. You cable vendor should have provided ...


5

This is a buffering issue. I had the same problem with a new Level 3 circuit. Their NID apparently has no measurable buffers, so feeding it at a rate 10x the circuit rate (1000 vs. 100) leads to all manner of poor performance. (They had the customer side set to auto (1000) but manually set the network side to 100. After a week of arguing, they set both ...


5

I assume you're talking about a structure like this: F | ISP | R S | | ===+===+===+=== | C If your router supports so-called "hairpin routing" (or "hairpin NAT"), then if you have this situation it will accept the request from client C and send it back out the LAN-side interface to server S. This has the ...


4

While UDP is not reliable, a lot of protocols use it as a base and add reliability at application level. You would determine what features you would need, and what you would implement: Connection handling (keep track of a connection) Sequencing (to rely on order of frames) Acknowledgement (to make sure all frames are received) Flow control (throttle the ...


4

User Teun Vink is very correct in his assertion that context is required, mainly in the platform that you are using and the configuration that the rate is being applied to, whether QoS or service utilization. Nevertheless, here is some basic information: CIR is your committed rate, which is the desired bandwidth of a service. It is not an indication of ...


4

@Machinarius, @Brett Lykins has pretty much answered your question but without mentioning any techniques specifically. If you wanted some further clarification then two main techniques very commonly used are shaping and policing depending on the service being contracted and the customer requirements. You will read about these in the links provided by @Brett ...


4

The DOCSIS Service Profile within the configuration downloaded when it registers on the network tells the modem (and headend) how much bandwidth an individual modem is allowed. Exactly how the modem (or headend) polices this is implementation specific.


4

First, there is a relationship between buffer size and latency only if there is congestion in the path (too many incoming packets on the input interface, or too many packets for the switching engine (ASIC/NPU/software/whatever), or too many packets in the output interface). On a "lightly loaded" switch, there is no relationship between buffer size and ...


4

If a network has a throughput bottleneck, does it matter whether the slow link comes before or after? No it doesn't, Only matters when destination is before or after that bottleneck link. What throughput speed will the red arrow have? 200 or 100? Yes it will be bottle necked at 100mbps. So bottleneck throughput speed would be of the lowest throughput ...


Only top voted, non community-wiki answers of a minimum length are eligible