I'm not sure if anyone else has come across this but I'm sure you are all aware that CentOS 5.3 default kernel comes with realtek support.
But the default realtek drivers have a bug in them when it slows down upload and downloads drastically and shows ping times above 2000+ ms in the same datacenter.
A fix for this is simply upgrading the kernel to the latest version available (2.6.29) with realtek driver support.
Say I have my own office to start building out my own "mini" data center.
If I wanted a lot of bandwidth for the most reasonable cost, how do one go about buying bandwidth from various providers? What is Metro-Ethernet vs. say a DS3? I know there is Fractional DS3 (6 to 45mb/s) and then Ethernet services up to GigE. I'd rather not mess with BGP right now either.
If I am looking at 15-20mb/Sec of bandwidth, what would be a good route to go? My budget is about $1500/month for bandwidth and IP Addresses (I need 2 Full Class-C IP Spaces). I also want to be able to turn up speed as needed without a lot of hassle.
Is this a reasonable budget? And where would be a good place to start? Just call up like Level 3 or am I too small potatoes to even deal with?
What low-cost Gigabit Ethernet switch would you recommend for providing residential IP services to an apartment complex that is already prewired for copper Ethernet? My only real requirement is that is it a managed switch supporting QoS, capping, and has fiber uplinks.
I now have to move on to the question of Ethernet ports in relations to rack mount servers. I have just about ever question answered now. Now my question is about the connection to the web my server will use. Is it better to have a single or a dual Ethernet post?
I see a lot of duals. Here are some of the samples I have read when building my server. 2 x Intel® 82546GB Gigabit Ethernet Ports Dual Onboard Gigabit Ethernet LAN Ports Dual Intel® 82541GI Single-port Gigabit Ethernet Controller Broadcom dual Gb NICs Intel dual Gb NICs Broadcom BCM5754, 10/100/1000Mbps
I have inbuilt Ethernet card RTL8139D , but when i install RHEL 4 or 5 , Fedora version the Ethernet card diver not installed or ethernet card not installed but when i install suse linux it working fine without any problem so can any one tell me how to install lan card on Fedora or RHEL5.I have driver in CD but it not working with it.
yesterday the dc tech logged-in by ssh (root) cause we couldn't connect to the ftp box we keep our backup's (the new box couldnt connect to the ftp got time out) They fixed it.
Today i got a time out and after 2 reset and restaring of services the server was back again.Then i got this messages and it still coming (3 hours now i get 1 similar message per 5 second )
Code: Nov 4 12:31:20 gandalf kernel: martian source 217.20.117.1 from 127.0.0.239, on dev eth0 Nov 4 12:31:20 gandalf kernel: ll header: ff:ff:ff:ff:ff:ff:00:1a:4b:3a:7e:23:08:06 Nov 4 12:31:21 gandalf kernel: martian source 217.20.117.1 from 127.0.0.244, on dev eth0 Nov 4 12:31:21 gandalf kernel: ll header: ff:ff:ff:ff:ff:ff:00:1a:4b:3a:7e:23:08:06 Nov 4 12:31:22 gandalf kernel: martian source 217.20.117.1 from 127.0.0.252, on dev eth0 Nov 4 12:31:22 gandalf kernel: ll header: ff:ff:ff:ff:ff:ff:00:18:71:80:07:86:08:06 Nov 4 12:31:22 gandalf kernel: martian source 217.20.117.1 from 127.0.0.244, on dev eth0 Nov 4 12:31:22 gandalf kernel: ll header: ff:ff:ff:ff:ff:ff:00:1a:4b:3a:7e:23:08:06 Also the firewall stopped the port scanning from a server within the same dc.
I checked it on google and the explanation's varies but most say that is a problem of the router/switch or similar.
We're ready to setup 3-4 42U racks for servers and are in need of choosing Ethernet switches. What do you guys use and why?
I'm looked at Cisco switches, but lost in their product forest. I'm looked at Express 500, 2960 and 3750-E models. Is there any more difference (in exception of stacking, cli and hot-swap fan/psu) that I need to consider? Prices differentiate too much.
I'm also looked at Linksys/D-Link business products and they seems to have the same features as Cisco Express series, but only 50% cheaper.
After a previous uplinks issue, i'm never able to hit above 4Mbps of download speed to my home's cable broadband anymore. I have no idea whether it's because my partner capped my bandwidth from Switch 1, or his usage has increased such that it's affecting my traffic too. I can't be too concerned about it with him, as he's paying the bulk of the bandwidth costs.
For the whole rack, i'm not able to hit above 5Mbps accumulatively, and my normal bandwidth usage is only about 1-2Mbps.
So i'm thinking of getting another line for bandwidth directly from the datacentre to my Switch 3, perhaps starting from just 1-2Mbps only.
How will things improve? Just a silly thought, whether my total available bandwidth can now be 5Mbps + 2Mbps = 7Mbps.
We have a few CPanel servers, all with multiple NIC's. (2 Each)
We want to setup Eth0 to be the managment interface (ie. 72.2.24.x) We want to setup Eth1 to be the Shared / Exclusive IP Nic (ie. 67.69.141.x)
Here is the issue, Eth1 is on a totally different Vlan than Eth0.. thus, isolating DDOS attacks, and any other type of traffic concerns for managment.. But, it also has a different G/W as Eth0.
my server has 2 nics, how do I view them and their usage uptime etc in debian stable I wanna see if they are being used randomly (usage on both), or if it hasn't been already setup and I need to do it (found a tutorial on how to do this, but dont' wanna fix it, if it ain't broken)
I am having a problem with YUM on Fedora 5, for a long time it has not worked saying there is nothing to update when I know there is, I have tried to correct this by following a forum post of someone having the same problem, now I have really messed things up...
When I type any yum command (yum, yum --version, yum check-update etc) I get the following error:
Traceback (most recent call last): File "/usr/bin/yum", line 28, in ? import yummain File "/usr/share/yum-cli/yummain.py", line 29, in ? from yum import logginglevels ImportError: cannot import name logginglevels
perhaps removing it and installing the latest version of Yum would help however tried that and it seems to make no difference, perhaps I am doing it wrong?
i have problem in my linux server that i have aserver with 5.2.4 apache and many scripts not working after that i upgraded to php 4.4.7 so all scipts all over my server are working but not loging with any user and pass just as u make refresh to the page this problem in all forums and scripts in my server i need quickly solution please my customers are confused
We recently setup a server with 4 gigs of RAM and installed Fedora core 7 32-bit version in it. After installing the OS, I have found that Fedora is able to detect only 2 GB and not 4 GB of RAM. I installed the kernel-PAE and kernel-PAE-devel modules and restart the server and made sure that the the OS with the PAE switch starts at boot time. However, the OS still does not detect the 4 GB RAM. Any idea what else can be done apart from installing the 64-bit OS in the system?
OS: FRESH install of Unix Fedora Core 6 Kernel Version: Kernel 2.6.18-1/2798.fc6 on an i686 Hardware Information: Brand new Dell SC1435 Server Software Version: MySQl 5 Control Panel: cPanel
A few things I did already..
#yum install mysql-server && service mysqld start Nothing to do mysqld: unrecognized service
#rpm -qa | grep mysql libdbi-dbd-mysql-0.8.1a-1.2.2 mysql-connector-odbc-3.51.12-2.2
#ps -ef | grep -i mysql root 15299 15114 0 16:55 pts/0 00:00:00 grep -i mysql
I don't see the mysql.sock file. What I can do to further troubleshoot this issue?
My server was just upgraded to FC6 and now I do not have pico for a editor. I found nano but there is problems. Screen does not refresh correctly and when I type in charaters sometimes extra charaters show up.
Are there any other screen editors built in to FC6 (not vi)
