The new setup - Memory

[AMH]

Hi,

Firstly the Java connection to the server is really horrible, however I see others having the same problem. My question is memory usage after the new setup. On the old system I used to average 5-600MB usage. Since the new system came along I now run at almost its maximum of 2GB. What has changed on the new setup that the memory usage is so high now?

Secondly I get memory drop offs with the server load which I dont understand in the grpahs. WHM shows that there server has not been rebooted for more than a day but these odd gaps persist. Is there something in the setup that it might not always record the memory and load usage? See B & C below.

Thirdly the memory graph appears upside down to me. What is the bandwidth usage - the white or the yellow as it says -1mb on the side to -3mb.

Cheers

 

[KH-Paul]

Hi AMH,

Many thanks for your feedback.
You don't have to use Java-based console connection - this is just a "backdoor" that can be used when you, for example, lock yourself out of ssh access one way or another. For normal shell access please continue to use your regular ssh client.

Such gaps on the graphs could happen when data collection cannot be completed for one reason or another. There is no need to panic when you see these gaps as they don't necessarily mean that something is wrong with the VPS or processes running inside the VPS. If you can PM me your VPS IP I'll check the logs in attempt to find out what prevented this data from being successfully collected there.

Memory usage graph current includes I/O cache as well which doesn't correctly represent the "real" usage. We'll correct this in in the future.

Your memory graph does not appear to be upside down, I guess you were referring to the traffic graph? If so - traffic graph shows incoming (external world -> your VPS) traffic above 0 line and outgoing (your VPS -> external world) below. This is done to be able to fully show both usages and to prevent them from being "covered" by each other.

 

[e-nes]

Hi AMH,
Memory usage graph current includes I/O cache as well which doesn't correctly represent the "real" usage. We'll correct this in in the future.

Hello Paul,
Approximately, when would we see this correction?

Until then, where is a good place to get a better representation of current VPS memory usage?

Thank you in advance for your assistance!