VPS Node, Load Hike

hidayat

Hello,

How to check if our vps is load hike in the host node? And how to fix it?

Thanks

Jack

What do you mean? Please explain in a bit more detail.

Do you mean ; you think your VPS is causing a load issue and are asking how to check if it is you?

Damian

Command: uptime

You'll see something like this:

 20:00:58 up 8 days, 20 min,  1 user,  load average: 0.08, 0.02, 0.01

Concern yourself with the first number of load average:. This number needs to be less than the number of cores available in your system. You can determine the number of cores available in your system with the following command: cat /proc/cpuinfo | grep MHz and counting the number of lines output.

There are other ways to see load, such as the sysstat package. Sysstat will also show you disk i/o and other factors.

@hidayat said: And how to fix it?

What virtualization tech?

hidayat

2 days ago, I am contact my VPS Support Provider. I am asking about the loading of website on my VPS is slowly (I mean, regading net speed). And they reply: "This was due to a load hike in the host node,and we rectified the issue. At present we are getting a good speed in your VPS ,please see the below given snippet."

hidayat

I am using an OpenVZ VPS from virpus

Jack

I guess that means there was a client causing an issue on the host node which caused an issue for you too.

Damian

What's the below given snippet?

24khost

The server as a whole was under a heavy load. Could be that you are on a node that is at capacity.

hidayat

==================================

[root@server /]# wget --output-document=/dev/null http://speedtest.wdc01.softlayer.com/downloads/test500.zip --2012-10-17 18:27:46-- http://speedtest.wdc01.softlayer.com/downloads/test500.zip Resolving speedtest.wdc01.softlayer.com... 208.43.102.250 Connecting to speedtest.wdc01.softlayer.com|208.43.102.250|:80... connected. HTTP request sent, awaiting response... 200 OK Length: 524288000 (500M) [application/zip] Saving to: `/dev/null'

100%[====================================>] 524,288,000 10.9M/s in 49s

2012-10-17 18:28:38 (10.3 MB/s) - `/dev/null' saved [524288000/524288000]

[root@server /]#

==================================

24khost

ahh the server might have been under a ddos and they null routed the ip.

hidayat

well, I was try to setup a firewall (like, iptables) and change shh port. But, not help out

24khost

no, your on a shared port so more like a ddos on another vps that was getting attacked would also slow down your connection.

hidayat

@24khost what do you mean shared port? I got 3 IPv4 Public from them and using it for my DNS

24khost

if your using a vps it is 1 server and you share a network port. Has nothing to do with your ip's each vps does not get it's own network cable.

hidayat

Oh, do you mean they was using load balancing methods and mixing on a hub? I am sorry, I still confused.

hidayat

And, how about a dedicate VPS? does its have own cable?

NateN34

Here is what a VPS is (best explanation I could find).

Basically, you have no control over what is happening with the speed....only the provider that owns the host node does.

vdnet

No provider is going to give you a dedicated link, especially with LEBs. This would require 20+ network cards and 20+ network cables per server, it just isn't going to happen. A gigabit port would help, but a heavy DDOS is likely to overload the router, not just the server.

FRCorey

You mean load spike.

concerto49

@vdnet said: This would require 20+ network cards and 20+ network cables per server, it just isn't going to happen. A gigabit port would help, but a heavy DDOS is likely to overload the router, not just the server.

Might as well get a dedicated server in this case? Whatever server that has this will cost more than the 20+ servers.