Tag Info

Hot answers tagged

4

Before you start building allowed IP range, identify the region. The list of S3 regions is available here. Let's assume you need an IP range for s3-us-west-2.amazonaws.com. You can get it by executing the following commands on a Linux system, or equivalent commands on other systems. Get IP address for the domain: $ host s3-us-west-2.amazonaws.com ...


4

BIND has a GeoIP plugin which, I believe, can allow you to change the response based on the user's location. An easier solution would be to use Amazon's Route53 DNS service which supports responding differently in different zones, record/set weighting, etc.


3

Does this mean that IBM owns everything in-between 9.0.0.0 and 9.255.255.255? Yes That's 3 octets a.k.a. 24 bits of freedom which means they own 2^24=16,777,216 addresses! Yep That seems like a lot for one company. There's no way they need that many, right? Correct


3

If you look at the iproute2 gitweb, you'll see it's showing the status of the RTN_ANYCAST bit set on the kernel routing structure. If you cross-reference that with the kernel source (rtnetlink.h) you'll see the following comment: RTN_ANYCAST, /* Accept locally as broadcast, but send as unicast */ If you check ...


2

192.168.1.3 to 192.168.2.254 aren't in the same /23. You need to either use 192.168.0.0 and 192.168.1.0 or .2.0 and .3.0 if you are set on the /23 subnet size. Better yet, switch to an even larger subnet; I'd recommend a /21 using 192.168.0.x to 192.168.7.x as that's approximately the largest a broadcast domain modern hardware can handle. You wont have to ...


2

If it's still in the DNS cache then it will be resolved to the internal ip address. Note that this would only occur if you didn't shut down and restart the client and connected from an external network. If you shut down and restarted the client before connecting from the external network then the DNS cache would be flushed and it would resolve to the ...


2

Another way to do this is to use SLP (Service Location Protocol). There is an open implementation here which supports most major platforms. http://www.openslp.org/ SLP or protocols like it or derived from it are how printers and such things get discovered on networks. The DNS method is certainly the simplest. But there is no announcement that the server ...


2

As suggested by others you can add static routes route add -host 10.10.10.5 dev eth1 and verfiy with route -n


2

A silent proxy may be indistinguishable from NAT. But some may add headers, that allow to identify that request comes through proxy.


2

On most of the networks I work with the router address is the first routable address (1, 129, and 193) in the subnet. This is by tradition, and not a requirement. Another option is to use the highest routeable address (126, 190, and 254) for the router. The lowest address in the each subnet (0, 128, and 192) is the network address and is not assigned. The ...


2

A few choices: Most easy one probably: NAT rule in POSTROUTING with the "owner" iptables module. If you run each process with its own username, you could have rules for each user to go out with a specific IP. Run each process inside a container such as LXC or Virtuozzo. Create a connection wrapper script for your applications to bind to an IP each and ...


2

The problem with ping and tracert is that they are not service diagnostic tools. Ping and tracert are only good as general connectivity tests if you know that the target should respond. Can ping tell me why my web site isn't available? No it can't. My ISP, router or web server may block ICMP but happily serve HTTP. Tracert is a better tool for detecting ...


1

You also can use network namespaces, if kernel allows. like described here http://blog.scottlowe.org/2013/09/04/introducing-linux-network-namespaces/


1

EDIT: From your additional comments, it sounds like you want to be investigating web analytics services. Don't reinvent the wheel. If you're not happy with the data you're receiving from existing analytics platforms, then it's likely because there's just not enough data available to reliably make the kinds of distinctions you're looking for. You may be ...


1

By default, non-transparent proxy inserts "Via:" field to HTTP request header. Of course, this is configurable option that can be disabled, for example: http://www.squid-cache.org/Doc/config/via/ Transparent proxy tries to be invisible, so cannot be detected: "A 'transparent proxy' is a proxy that does not modify the request or response beyond what ...


1

Speaking for myself, we acquired a /26 block because 'we needed it' last year. Then we came to our senses and re-architected our SaaS product. Still have the block though - no reason to get rid of it - and frankly it's easier to have many. My last employer - not a small company had a handful and it was a constant PITA for the network guys having to juggle ...


1

Those addresses are network addresses, which identify the specific network, and cannot be assigned to a host, which is what a router is. A router's interface must have a host address in order to identify that it is a member of that network and to know what interface to route traffic to.


1

IGMP is only used within a broadcast domain, eg within the switch or switches. Between domains, anything that gets routed, the multicast group information is communicated by another protocol - almost always this protocol is PIM. I'm not sure I understand your question though. The client joins the multicast group, the serve don't force it into a group. If ...


1

First, you run a traceroute. $ traceroute 65.49.51.227 traceroute to 65.49.51.227 (65.49.51.227), 30 hops max, 60 byte packets 1 172.25.50.1 (172.25.50.1) 0.364 ms 0.473 ms 0.609 ms 2 172.25.49.1 (172.25.49.1) 3.722 ms 3.828 ms 4.962 ms 3 73.195.236.1 (73.195.236.1) 33.326 ms 34.256 ms 34.357 ms 4 ...


1

As @Michael Hampton said, traceroute is your best friend. I am able to resolve and access the website just fine. Are you able to at least ping it? If you can't ping it then you are probably blocked somehow. Is there a firewall running on that server? Can you use another IP to reach it?


1

Hurricane's webhosting servers don't respond to traceroutes. The same goes for a lot of intermediate routers of a lot of ISP's. When submitting issues like this, always run ping as well as traceroute, and include both results. Pathping, mtr, and WinMTR are also helpful tools.



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