Memory Load Average

[Flava]

Hi,

Is it possible to do memory load average? I know we can do this for CPU, i.e we can configure nagios to alert if CPU is 80% load on average of 30mins and so on, can we do the same for memory for servers with fluctuating memory threshold. I dont want to get an alert every time memory consumption increases. I have already already increased the threshold, it is configured at hard size (256MB) rather than using percentage.

[mguthrie]

To my knowledge all of the memory plugins look at current usage. However, what I would recommend is setting your retry_interval and max_check_attempts to a value that would let you know that the memory usage is high over a period of time. So for example, lets say I only want to get an alert if the memory usage is above my threshold for more than 30 minutes.

Code: Select all

check_interval 5
retry_interval 3
max_check_attempts 10

What will happen is that if Nagios detects a problem, it will retry every 3 minutes up to 10 times before sending an alert. If the memory threshold is exceeded for a full 30 minutes, then I'll get the alert.

[gshergill]

Hi mguthrie,

Thanks for that post, I didn't realise Nagios had that kind of functionality.

In the example you gave, the alert happens if the memory usage is above the threshold for more than 30 minutes, I assume that the threshold is referring to the warning value?
If during the 30 minutes it records the usage as being above the critical value, will an alert be immediately sent out? Or would it wait the full 30 minutes?

Thank you.

Kind Regards,

gshergill

[mguthrie]

It will send out the alert after the if the service is in a "Non-OK" state after X amount of retries. X being the max_check_attempts.

[gshergill]

Hi mguthrie,

Thanks for clearing that up.

Kind Regards,

gshergill