Tedd Johnson's questions

I have a VPS server running Windows 2012 R2 that is used with a piece of software that makes roughly 15,000 connections in a minute. I noticed some lag and other issues with the VPS, and decided to upgrade to a Dedicated server.

VPS Specs: "4 vCores 3.1 GHz, 8GB Ram, SSD, 100mbit connection" Dedicated Specs: "Xeon E5-1620 v2 3.7GHz, 64GB RAM, 1gbit connection"

I installed Windows 2012 R2 on it, and noticed a dramatic speed decrease in the software. It was hitting only about 5,000 connections per minute. I thought it might be because this was on a HDD instead of an SSD, so I setup a RAMDisk. There was zero improvement in the performance. I started to look into TCP limitations, thinking the VPS container that was automatically setup for me might be different than my settings. I couldn't find much. I added a ton of options to the TCP Parameters in the registry, and still see no difference.

[HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Services\Tcpip\Parameters] "TcpTimedWaitDelay"=dword:0000001e "MaxUserPort"=dword:0000fffe "TcpNumConnections"=dword:00fffffe "TcpMaxDataRetransmissions"=dword:00000005

The software connects to a single IP address, across multiple ports (~100 ports).

I'm unsure of where to look now. All speed-tests show the dedicated server is significantly faster than the VPS, but the VPS, with a slower connection and less powerful hardware, can reach faster connections per minute.

Any direction to where I can look further would be appreciated.

I have a server that is not a virtual machine, and has plenty of RAM (32GB) and its using less than 6, but I constantly get this error when running some programs.

root@DAL1-727]# free -m
bash: fork: Cannot allocate memory

if I then kill something benign (like an extra bash session I had open in screen) I can get it to let me do things:

[root@DAL1-727]# free -m
             total       used       free     shared    buffers     cached
Mem:         32067       5015      27052          0         55       1293
-/+ buffers/cache:       3666      28401
Swap:        16095          0      16095

I'm opening a lot of files and apache is probably the cause. I've tried setting my file limits higher, but this still happens.

ulimit -n 1048576

output of /etc/security/limits.conf

soft nproc 65535
hard nproc 65535
soft nofile 65535
hard nofile 65535
apache hard nofile 64000

What else could this be?

I've tried searching for this answer, but haven't found anything elegant.

I have numerous servers in a colo that is in another state. I need to find a way to check that the servers have RAID-1 on them, so that I can determine if they were setup correctly by my colo.

df -h shows: Filesystem Size Used Avail Use% Mounted on /dev/mapper/VolGroup00-LogVol00 442G 1.5G 418G 1% / /dev/sda1 99M 19M 75M 20% /boot tmpfs 4.0G 0 4.0G 0% /dev/shm

however as CentOS uses LVM by default, this doesn't indicate if a RAID-1 is present.

it is supposed to be a software raid, so I'm pretty sure there should be a way to check.

Thanks

I have a system that receives many different DNS requests every day from unauthorized users. What I need is a method to accept all DNS requests made to to this system and give back them back the same IP. Think of this as a gateway situation, where regardless of their DNS request, it will hand back the same IP.

Is it possible to do this through BIND? or is there a better more elegant solution?