I am going to collocate two servers in a datacenter and I wish to use my own IP addresses (as I’m member of the RIPE).
I wish to get a block for this, what is the minimum size of a PI block to make my collocation provider able to route them?
Is it /24 PI minimum, or is it also possible to just cut a /28 for this purpose?
I have been trying to figure out how to allow a range of IPs with APF. I finally came across the CIDR page on wikipedia and read it though, and now I get it. I just want to confirm one thing before I implement this.
I run a gameserver where I want to allow everyone in the world to connect on a certain port, so can I use this entry?
Code:
tcp:in:d=27015:s=192.168.0.0/0
From what I understand, using the /0 CIDR suffix basically means "all IPs". So technically I could use ANY IP address for the source with the /0 suffix and it will work right?
We are considering a VPS and was wondering what minimum RAM we should have. For cPanel, I ahve read one should start with 512 unless the VPS is fully tuned. Is there a general of thumb for Linux powered plesk based VPS's?
View 4 Replies View RelatedHas anyone put together a list of minimum packages required for a LAMP box on RHEL4?
After doing a so-called "minimal" server install, there are several packages I uninstall during the hardening process (anacron at autofs cups* gpm irda-utils isdn4k-utils nfs* pcmcia-cs portmap redhat-lsb xinetd yp-tools ypbind). While deploying a new box recently, I got to thinking about how many other packages included in the so-called "minimal" install aren't really necessary on a typical LAMP server. Obviously the answer is "it depends", but... what else are folks uninstalling?
reliable host offering around 20-30gb storage and max bandwidth..
View 11 Replies View RelatedHow much would the minimum requirement for guaranteed RAM be on a VPS with cPanel?
View 8 Replies View RelatedIm trying to host a video sharing site. I have been through 2 hosts that told me they support the files that i need (i.e. flvtool2, mencoder, mplayer, ffmpeg, etc.) Come to find out that they dont. Now i am in need of a new host that supports this. I have researched different ones but to make a decision i need this answer:
For a video sharing site just starting out what is the recommended bandwidth, data transfer, and disk space to run this site? I just want to start small and i will expand if the growth permits it. I just dont want to jump in feet first not knowing the outcome.
i am hosting my site in shared hosting. My totall visitor 5000. All i want to know when i should upgrade from shared to vps 512mb CPU Speed 2.0Ghz? Means how many visitor online i need to upgrade to vps? Please help. My current shared hosting plan include 20gb space and 400gb bandwith. Currently 5000 user using 4 gb space and 120 gb bandwith.
View 10 Replies View Relatedif there is any way to forward an external IP to an internal subnet without NAT.
I have a server that is configured with a 10.0.100.101 IP and the L3 switch doesn't support NAT, so I can't get on it right now without manually changing the IP on the NIC to a public IP address.
I have a linux router with 2 external and 2 internal ports.
Each port needs to route traffic to one of the internal ports, and the internal traffic between the 2 internal ports should not go out the external ports.
The IPS on the internal networks are global. ie. no NAT required.
I think what I need is this..
$ext_net1 = external nework IP/MASK 1
$EXT_IP1 = ip of external interface 1
$ext_net2 = external nework IP/MASK 2
$EXT_IP2 = ip of external interface 2
$int_net1 = internal network IP/MASK 1
$int_net2 = internal network IP/MASK 2
ip route add $ext_net1 dev eth0 src $EXT_IP1 table 1
ip route add default via $ext1_gw table 1
ip route add $int_net1 dev eth1
ip route add $ext_net2 dev eth2 src $EXT_IP2 table 2
ip route add default via $ext2_gw table 2
ip route add $int_net1 dev eth3
ip rule add from $int_net1 table 1
ip rule add from $int_net2 table 2
I have a Webmux load balancer and behind that a Cisco Pix. Behind that I have several servers. The Webmux and Cisco Pix do double NAT so his servers have public IPs.
The problem is that I've added a 4th server, I added it to the Webmux and it's get NATted to an 192.168.x.x IP. Now I just need to add it to the Cisco Pix, natting it back to the real IP BUT the Pix can only have one IP on its inside interface and the Server IP is not on the same subnet as that IP.
So when I try to add the real IP it asks me how to route it....
I'm trying to implement VLANs on my network and can't get connectivity to host servers. Here's how the network is configured. Pardon the bad ascii diagram.
In this example my upstream is providing two subnets:
111.111.111.16/28 (I'm using an IP from this subnet to manage the 3550)
222.222.222.16/29
I am attempting subdivide the /29 into two /30's in order to place a server into it's own /30 subnet & VLAN ............
I have a server that has multiple IPs, one of which I'm using for a VM that is bridged.
The issue is, internally, that IP is trying to point to itself rather then the bridged nic (which is technically a whole other server plugged into the same switch, logically).
I think I know why, I just don't know how to fix it. This is the config file for the ranges:
Quote:
# Intel Corporation 82546EB Gigabit Ethernet Controller (Copper)
DEVICE=eth1
BOOTPROTO=static
DHCPCLASS=
HWADDR=....
ONBOOT=yes
IPADDR_START=....243
IPADDR_END=....254
CLONENUM_START=0
GATEWAY=....241
NETMASK=255.255.255.240
NO_ALIASROUTING=yes
(edited a few things out just in case)
Basically, there's a start and an end, is there a way to exclude an IP?
to expand our existing DNS setup with nodes in North America and Asia.
Therefore, we are searching ISPs that can provide dedicated servers and route an existing (RIPE PI) IP range to that server which will be anycastet for DNS service?
What company would be able to provide that service?
I put together a router running Zebra(yes, I know, should have used Quagga) with a few public ip addresses taking in a full BGP table.
There is a Win2k3 server behind the router running routing and remote access for VPN clients to connect to. Our team's project was to get the win2k3 server VPN clients out onto the public internet with public IP addresses.
I installed another NIC card onto the Win2k3 server and connected it with the router, and assigned the router and the server a private IP address. Both are pingable from both devices.
I then had a VPN client connect in, RRAS assigned the client a public IP address, the router was able to ping the VPN client and so was the Win2k3 server.
I tried pinging the VPN client from another machine on the network with a default gateway pointed toward the router, and there is no response.
Is there something I don't know about with Zebra and Redhat?
I'm experiencing some odd issues, I have a cpanel setup, however on port 2086 the server is currently listening however on port 80 it fails to listen. Apache is running and no errors appear in the errors log.
Running ifconfig shows that there are errors and dropped packets.
I was changing ip routes earlier that day however all seems fine...
Oddly I can ping internally on the network and noticed a number of other servers in the broadcast range. These respond fine, however pinging google or outside the data center fails.
ping google.com
ping: unknown host google.com
HTML Code:
eth0 Link encap:Ethernet HWaddr 00:14:85:3D:A2:20
inet addr:122.252.4.* Bcast:122.252.4.255 Mask:255.255.255.0
inet6 addr: fe80::214:85ff:fe3d:a220/64 Scope:Link
UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
RX packets:289198921 errors:4 dropped:182 overruns:0 frame:3
TX packets:230175646 errors:19 dropped:0 overruns:0 carrier:19
collisions:8927682 txqueuelen:10
RX bytes:3521641159 (3.2 GiB) TX bytes:2563591520 (2.3 GiB)
Base address:0x2400 Memory:dd100000-dd120000
lo Link encap:Local Loopback
inet addr:127.0.0.1 Mask:255.0.0.0
inet6 addr: ::1/128 Scope:Host
UP LOOPBACK RUNNING MTU:16436 Metric:1
RX packets:127443475 errors:0 dropped:0 overruns:0 frame:0
TX packets:127443475 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:0
RX bytes:1102069037 (1.0 GiB) TX bytes:1102069037 (1.0 GiB)
I have the following problem with a CentOS server:
The main IP of the server is yyy.zzz.www.qqq
We've just purchased 3 additional IPs: aaa.bbb.ccc.100, aaa.bbb.ccc.101, aaa.bbb.ccc.102.
First, all outgoing traffic used aaa.bbb.ccc.100, but after deleting the gateways from the additional IPs it seemed to work fine, until we found out the following:
Now all traffic to aaa.bbb.ccc.XXX uses aaa.bbb.ccc.100 as outgoing IP.
What command would change this to use our main IP?
I am setting up a small ccna lab and i have RIP working and i can ping my lan from both routers, but only certain hosts on the lan from the one router the setup is
LAN (192.168.1.0/255.255.255.0)
|
/
router 1 E0 192.168.1.45
Serial0 10.10.10.1
|
/
Serial0 10.10.10.2 (of router2)
|
/
E0 192.168.3.250
E1 192.168.2.250
Lo 192.168.5.4
I can ping 192.168.1.102 from router 2 and 192.168.1.45 but no not 192.168.1.201 ... or 192.168.1.1
also i can ping 192.168.5.4 from 192.168.1.102 which is a linux box and an ip route to tell it that 192.168.5.0 can be gotten from 192.168.1.45
Can someone recommend me good (and cheapest) routing or switching gear for the following scenario?
Multiple 1Gbit links, possibility to have a single 10GE link
Sustained 2Gbps of traffic, 4Gbps of peak traffic (streaming media)
I've been doing some traceroutes between Chicago and Dallas. Tracing from Chicago -> Dallas, I go through Denver almost 100% of the time. Tracing from Dallas -> Chicago, I go through Denver or Atlanta before routing to Chicago.
Is this normal? Looking at the Level 3 network map there seems to be several, much shorter routes.
I run a game server on The Planet, and lot of people have huge routing issues where their route randomly changes, and when it does, they'll get horrible packet loss and lag. It's totally random, one day it may happen to me, while it's not happening to someone else, then it will switch. But it's definately the host and not our home connections as it affects about half the server at any given time, it just picks different people.
Just wondering if anyone who uses The Planet has had issues like this? I pretty much debugged everything and tried everything to no avail and of course their support just said it's not at their end (all isps of any type say that regardless of the situation).
This is how a typical trace route would look like:
Code:
3 9 ms 9 ms 19 ms GE-2-1-ur01.N3Alpharetta.ga.atlanta.comcast.net
[68.86.110.17]
4 8 ms 12 ms 7 ms 68.86.106.133
5 8 ms 14 ms 13 ms 68.86.106.129
6 9 ms 8 ms 19 ms 68.86.106.125
7 9 ms 7 ms 8 ms 68.86.106.13
8 22 ms 7 ms 8 ms 68.86.106.9
9 11 ms 11 ms 8 ms 68.86.90.121
10 29 ms 21 ms 39 ms te-0-7-0-0-cr01.nashville.tn.ibone.comcast.net [
68.86.84.65]
11 31 ms 66 ms 30 ms te-0-0-0-4-cr01.chicago.il.ibone.comcast.net [68
.86.84.77]
12 50 ms 41 ms 56 ms 68.86.84.17
13 44 ms 45 ms 53 ms 68.86.85.38
14 53 ms 49 ms 50 ms 68.86.85.45
15 49 ms 51 ms 59 ms te-7-3.car1.Washington1.Level3.net [63.210.62.57
]
16 57 ms 53 ms 54 ms ae-32-52.ebr2.Washington1.Level3.net [4.68.121.6
2]
17 79 ms 93 ms 86 ms ae-2.ebr2.Chicago1.Level3.net [4.69.132.69]
18 * * 103 ms ae-1-100.ebr1.Chicago1.Level3.net [4.69.132.41]
19 115 ms 110 ms 126 ms ae-3.ebr2.Denver1.Level3.net [4.69.132.61]
20 125 ms 178 ms 126 ms ae-1-100.ebr1.Denver1.Level3.net [4.69.132.37]
21 132 ms 128 ms * ae-2.ebr1.Dallas1.Level3.net [4.69.132.106]
22 141 ms 130 ms 131 ms ae-14-55.car4.Dallas1.Level3.net [4.68.122.144]
23 130 ms 140 ms 129 ms THE-PLANET.car4.Dallas1.Level3.net [4.71.122.2]
24 130 ms 141 ms 130 ms te7-2.dsr02.dllstx3.theplanet.com [70.87.253.26]
25 * 130 ms 134 ms vl42.dsr02.dllstx4.theplanet.com [70.85.127.91]
26 135 ms 138 ms * gi1-0-1.car11.dllstx4.theplanet.com [67.19.255.4
2]
27 127 ms 135 ms 133 ms a.c4.1343.static.theplanet.com [67.19.196.10]
Another:
Code:
4 209.226.50.77 (209.226.50.77) 49.145 ms 46.724 ms 47.563 ms
5 142.46.7.1 (142.46.7.1) 55.852 ms 56.377 ms 55.110 ms
6 142.46.128.53 (142.46.128.53) 59.420 ms 56.865 ms 59.141 ms
7 142.46.128.5 (142.46.128.5) 59.277 ms 61.681 ms 59.702 ms
8 ge-1-1-0.ar1.YYZ1.gblx.net (64.212.16.81) 59.951 ms 58.555 ms 58.397 ms
9 por4-0-0-10G.ar2.DAL2.gblx.net (67.17.105.38) 95.604 ms 98.524 ms 97.206 ms
10 The-Planet.GigabitEthernet7-3.ar2.DAL2.gblx.net (64.208.170.198) 252.656 ms 251.881 ms 251.271 ms
11 te7-2.dsr01.dllstx3.theplanet.com (70.87.253.10) 253.416 ms te9-2.dsr02.dllstx3.theplanet.com (70.87.253.30) 252.040 ms te7-2.dsr02.dllstx3.theplanet.com (70.87.253.26) 251.873 ms
12 vl41.dsr01.dllstx4.theplanet.com (70.85.127.83) 255.683 ms vl42.dsr02.dllstx4.theplanet.com (70.85.127.91) 257.144 ms vl41.dsr01.dllstx4.theplanet.com (70.85.127.83) 263.597 ms
13 gi1-0-1.car11.dllstx4.theplanet.com (67.19.255.42) 259.076 ms gi1-0-2.car11.dllstx4.theplanet.com (67.19.255.170) 262.143 ms gi1-0-1.car11.dllstx4.theplanet.com (67.19.255.42) 263.775 ms
14 a.c4.1343.static.theplanet.com (67.19.196.10) 264.516 ms 265.046 ms 264.407 ms
-bash-3.1$
Actually if anyone is interested in looking this more I posted a thread here. But not needed. Just want to know if anyone else has had issues like this with The Planet. The only thing I can think of right now is switching hosts, but thats an expensive process as for the transition time I'll be paying for two hosts.
Our colo has two carriers, call them A and B. I have discovered the colo provider is round-robining traffic out it's two carriers on a per-packet basis, not per flow.
Assume we want to reach destination IP a.b.c.d.
%> traceroute -q5 a.b.c.d
Results show that at the hop leaving the colo's border router, some packets transit Carrier A and some Carrier B, to the same destination IP, during the same traceroute.
Is this a routing Best Practice, or am I correct in thinking this is the Lazy Man's way of load balancing across multiple circuits, multiple carriers? BGP route selection does not seem to apply here (i.e., either Carrier A or Carrier B but not both at the same time).
if using the Internap FCP technology to optimize the routing. I want a feedback on it, since I want to deploy this solution for have a better traffict routing.
Also, is anyone using avaya? i have looked in their website, but I have no information about their routing optimizer. Basically I want to go beyond of normal BGP since I will be deploying VoIP services soon.
We have a weird CentOS routing problem:
We need traffic to a certain subnet to go out via a second interface IP, rather than the main IP.
I.E, eth0 has IP x.x.x.x and eth0:1 has IP x.x.x.y (on the same subnet). I want traffic to z.z.z.z to go with a source of x.x.x.y rather than x.x.x.x like all the other traffic.
However I add the route and specifiy the device eth0:1 it accepts it but it goes into the routing table as eth0, whether I do it through network-scripts/route-eth0:1 or route add -host z.z.z.z gw a.b.c.d dev eth0:1.
When I ping with the -I command for eth0:1 it works, so the idea works fine, I just don't want to have to specify the interface in the application, but to do it within the routing table.
This is on CentOS 5 under Xen but I've tested on CentOS 4 under Virtuozzo too and it's the same.
I have an exchange server with the webmin interface activated. I have port 80 forwarded on the router to this server for the webmin. Management wants to upgrade data service and move website server in-house.
So this is how I'll need to set up 2 websites on two servers for the same IP.
Server 1: Windows. accepts domain mail.domain.com
Server 2: Linux. accepts www.domain.com
I'm figuring I'll need to make changes at the router level, and I have a decent cisco router. What do I need to do? Point me in the right direction for googling.
Would it be easier to use 2 IPs? Both would come over the same line, how would I handle that on the router level?
we're making a major move to offer some new services, etc, so we're going to have a couple racks in a datacenter, and will be running our own network. Currently I've purchased a 2821 router and a 3750G-24-T-S switch. I've got an ethernet drop from Cogent and we're set up for static routing.
I'm waiting on pricing from Sprint, Verizon and Level3 for a second ethernet line, and I'd like to multihome between them. It's been a while since I've done much cisco configuration, so I'm a little rusty. I know I'll have to run BGP between my routers and the provider, and IBGP between the two routers.
I'm looking for suggestions to run internally. We'll be offering standard shared hosting, a vps/cloud solution, dedicated servers and standard colocation.
Some people have suggested to run each of my different products on a separate VLAN, but I'm worried that would be wasting IP addresses, since one would be needed for the default gateway of each. For instance if give a /29 with my dedicated servers or my higher VPS plans, then it uses up 8 ip addresses, but the customer only gets 6. That seems right, unless I'm thinking about this incorrectly.
So as for my main question... If I did it that way, would I be ok to just run EIGRP on the 3750 to route between vlans, and then redistribute a default route from BGP into EIGRP?
I have VPS and on my WHM (Visual Mail Troubleshooter) i make test and show me Mail Routing Error.
View 4 Replies View RelatedI have spent the last few days setting up a low-end VPS server as a VPN host, using OpenVPN on CentOS 5.
I've got everything set up, but one last (and most critical) component is still not working correctly.
Basically what I need, is that once users are on the VPN, they should be able to browse the internet through the VPN under that server's IP address rather than their own dynamic address. I was told that this VPN set up was the way to do it. However, right now when I connect into my VPN I can browse that specific server, but cannot access any other websites at all.
If it would help to see my config files, please let me know and I'll post.
I'm really itching to have this up and running,
business is just getting out of hand. I applied several weeks ago for Whitelist status, and my issues finally went away for a little over a week (though I never received a response to my Postmaster requests). But then today -bam- 100% deferrals for going on 18 hours now, not a single message has gone through. And naturally no two Yahoo servers give me the same error message.
So...
At this point I'm ready to contract out my Outbound mail to Yahoo through a whitelisted 3rd party until I can get this resolved on my end. Would this be reasonable? Is anyone else doing this? I worked with an outsourced SMTP provider in another life for an internal company mailing list with good success.
On the same server, can I use different outing IP for each domains?
They assign 5 Ips to me. but I don't know how to do so.
Panel: Webmin