25
Yes, this does happen quite a lot, and it is called private peering. It has some benefits over peering over an IXP:
dedicated bandwidth, you can be sure you can use the full capacity of the interconnecting link for traffic to and from the other ISP
no dependency on the IXP, an IXP connects two ISPs on their switch(es), you're not suffering from any outages ...
22
There are a number of things ISP's do:
drag bundles of fibers across continents. Since this is very costly, only a small number of very large companies do this and many ISP's rent fiber-pairs from these companies
rent capacity (a wavelength, VLAN, MPLS circuit, etc) between to an IXP from a company who owns (or rents) fibers. Since the capacity of fibers is ...
14
The RIRs assign addressing to the ISPs. An ISP not following the rules will quickly find itself ostracized and cut off from the rest of the Internet.
IANA owns the addressing and assigns each of the five RIRs blocks of addresses, and the RIRs assign blocks of addresses from those to businesses that can prove they need them and are willing to pay for them. ...
12
EDIT: I forgot to mention this - if you're interested, there have been books written about this topic. I highly recommend Bill Norton's The Internet Peering Playbook. Available in print or digital copies - it's pretty much the de facto text on this kind of stuff.
DISCLAIMER: The examples used here are hypothetical only - I have zero insight into the ...
10
IANA is responsible for the assignment of IP-blocks and AS-numbers to regional internet registries (RIRs). There are 5 RIR's:
RIPE (Europe, Middle East, Central Asia)
AfriNIC (Africa)
APNIC (Asia-Pacific)
ARIN (North America)
LACNIC (Latin America)
Each of these RIRs is responsible for assignments to ISPs in the region they service. They all have ...
10
You are pretty accurate.
IANA is the Internet Assigned Numbers Authority. Its function is to make sure that numbers that need to be unique are actually unique. That includes protocol values, IPv4 addresses, IPv6 addresses and AS numbers.
For IPv4, IPv6 and ASNs we have a distributed system: IANA doesn't assign those numbers directly to the users. They ...
10
Most likely they mean a wavelength on a DWDM path. With DWDM, a fiber operator can use one single pair of fibers to carry a number of different colors (wavelengths) of light using multiplexing. Each of these colors can then be used to transmit its own signal, for example 10Gbps ethernet. This way, they are able to use a fiberpair more efficiently.
8
1.Would it be better to setup traffic policing?
You're using shaping, which is better in this case. The important thing is to buffer your traffic before sending it to your internet provider. Consider this:
Your ISP asks you to set your physical interface to 100Mbps
Your service from the ISP is 30Mbps
What happens if you're already sending 30Mbps of ...
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
Based on your edit, it sounds like you want anycast. This is a common among large, global companies. You advertise the same network from multiple places, and routing will take traffic to the closest (from a network perspective) site advertising that network. When one site goes down, routing will automatically take traffic to the next closest location.
...
7
Even if you could still get PI IPv4 addresses in Asia: if your ISPs don't want to route your IP addresses then there is nothing you can do. Tunnels and LISP could solve some of your problems (I use LISP here), but you already stated that this is not available in your region.
BGP is the protocol that is used to route your IP addresses from an AS. You need ...
7
Who pays for traffic when it transits multiple providers?
Ultimately, you do.
John Jensen did a fantastic job explaining how this works overall, but I suspect a diagram could help. I will borrow the example providers he used.
Let's assume you have a web server colo'd in an HE.net facility; assume you pay HE.net $500/month for a 100Mbps co-colocation. ...
7
At this moment, there's not that much that ISP's can do to prevent these kind of hijacks except, as you said, doing proper filtering of received and announced prefixes. Unfortunately many ISP's still don't do that. RPKI may help as well but has some other drawbacks and isn't widely implemented either.
I wouldn't expect on-net traffic for ISP-A to be ...
7
The only answer we can give you is 'talk to your ISP'. Remotely triggered blackholing (RTBH) is an often used technique, but we can't tell if your ISP supports this.
When using RTBH, you can announce to your upstream which target IP's should be blackholed, either over a dedicated BGP sessions or by tagging target prefixes with a specific community.
7
What it does is simplify your network administration, particularly for large high-traffic networks.
Running dual stack means administering two seperate IP allocations for every network. Running IPv6 with NAT64 means running only one.
Dual stack networks can be a pain to debug, it can cause a lot of confusion if one protocol is working, but the other is not.
...
6
I have no practical experience in running CoS/DSCP in ISP networks, but a logical explanation would be that it doesn't make sense to just change fields in a header. Even if I as an intermediate ISP wouldn't be interested in these fields, the receiver or some other intermediate party could be. Resetting DSCP fields just because I'm not using them sounds like ...
6
Most transit provider will refuse to announce tough BGP a prefix longer than /24. (I.E. /25 for example)
It's related to the size of the routing tables in the routers.
You have to understand that the prefix you announce will be propagated to almost all routers on the Internet, and that's a lot.
Each prefix announced take some memory in each router, and ...
6
Most ISPs have to buy service from other ISPs to be able to reach every part of the internet. That is called "transit". If you start a small ISP you usually buy transit from one or more bigger ISPs. They will have connectivity to the whole internet, and they will sell that connectivity to you.
You can also set up direct connections to other networks. This ...
6
In general, traffic to addresses within your network should never go via your upstream ISP. Routers in your network should have a route for the every IP prefix in your network and exchange those via iBGP.
In addition, it's good practice to never accept routes towards (more specifics) of your own IP prefixes from your upstreams, since you should always be ...
6
These are not different types of routers, they are different roles. It usually depends on where, topologically, the device placed in the network. Edge routers, for example, are placed at the network edge, while a core router is, well..., in the core.
Routers all do the same thing: they forward packets based on layer 3 information. they might have ...
6
My understanding is that IXPs provide the primary means of ISPs to connect with each other
I think that is a massive oversimplification.
In general connections between autonomous systems (this includes ISPs but also other major networks) can be split into two main catogories (there are also intermediate cases). In a "transit" connection a customer AS pays ...
6
While it might not be possible to solve your narrow question about IP addresses without provider-independent addressing and control of your own routing, it is certainly possible to do something the wider problem of losing live chat connections because your IP address changes.
1. Better line A higher quality line might go down infrequently enough this ...
5
Routers are layer-3 devices. They communicate with other layer-3 devices (routers, NAT devices, layer-3 firewalls etc.) over layer-2 infrastructure (T1, ethernet etc.).
ISP networks can be quite complex. Here in The Netherlands we have many DSL connections like this:
Customer edge router (can be a NAT device) connects using PPPoA (over ATM/ADSL) to a DSLAM
...
5
The first diagram will not work, and in fact may result in damage to your router. The AUX port is a regular RS232 serial port, which is completely incompatible with ADSL. Don't do it.
The second diagram should work, provided your D-Link modem can operate in bridged mode.
5
You need to think of the difference between IP Range "Announcement" vs Origin AS. So from the perspective of what Hurricane Electrics AS6939 sees announced from its Peers. If you look at the Whois Tab information you will see the "Origin" AS.
To really answer your question though, how do you find out Who really owns WHAT. The arin whois lookup will tell ...
5
There are a number of possibilities here, not limited to, but including:
The CPE's CPU is maxing out. (CPE=Customer Premesis Device)
Check the specifications for the router you are using to make sure it can support the level of traffic you're trying to push. Try and graph the CPE's CPU if you are unsure.
Bufferbloat in your ISP
If your ISP has configured ...
5
In my opinion, you've set things up "wrong". (and so has the ISP)
Never, EVER, trust what the ISP puts on your link. If you're running a metro-e interface into a switch, everything possible should be shut off on that port -- VTP, CDP, LLDP, spanning-tree, etc. The ISP isn't part of your "enterprise", so there's no reason to exchange such information with ...
5
To start, the reference you are using is incomplete. Here is another one, that is also incomplete (lists none in Milwaukee and only one in the Chicago area). Both are missing one of the largest IXPs in Chicago.
But more importantly, these are only the points where one network will peer with another network. This doesn't account for the breadth or scope of a ...
5
The DNS is issued by default from your local host, not from the proxy, when using SOCKS (probably because SOCKS4 didn't support other way).
Make sure you have checked the checkbox "Remote DNS" in the Proxy configuration of the Firefox:
5
I think you misunderstood how a router process a packet, thus coming with a solution that is not at all appropriate for your needs.
Why?
Let say computer A has the following configuration:
mac address 00:53:BA:12:17:19
IP address 192.168.0.7
subnet mask of 255.255.255.0
default gateway 192.168.0.1
A send a packet to the internet host www.example.com ...
Only top voted, non community-wiki answers of a minimum length are eligible
