Page 1 of 1

Memory usage showing wrong Alert

Posted: Wed Mar 20, 2019 1:24 pm
by manimurugesan
Hello,

We are facing issue for memory usage for AIX servers . Please see the below status information

WARNING - 6306 / 57344 MB (10%) Free Memory, In Use: 51037 MB, Pinned: 5454 MB

Current threshold limit warn 10 and critical 5.

Total size 57344 MB , if it is 10 % free means 5734 MB but now it's showing 6306 but i can see the warning alerts in nagios console .

I mean free space is showing more than 10 % that is 6306 MB as per the Current threshold limit less than 5734 MB only it should show warning alert .
Please find the below output on aix server

#vmstat 2 5

System configuration: lcpu=16 mem=57344MB ent=2.00

kthr memory page faults cpu
----- ----------- ------------------------ ------------ -----------------------
r b avm fre re pi po fr sr cy in sy cs us sy id wa pc ec
0 0 3064498 1553275 0 0 0 0 0 0 103 14668 1402 8 1 92 0 0.28 14.2
1 0 3064479 1553293 0 0 0 0 0 0 67 4523 998 2 0 98 0 0.08 4.1
2 0 3064498 1553274 0 0 0 0 0 0 42 7381 996 3 1 96 0 0.13 6.4
4 0 3064481 1553291 0 0 0 0 0 0 71 12911 1005 3 1 96 0 0.14 6.8
1 0 3064479 1553293 0 0 0 0 0 0 964 13210 2883 19 2 80 0 0.70 34.8


Please find the attachment for SVMON command output.

Could you please help me understand this and please help us to resolve this issue ?

Re: Memory usage showing wrong Alert

Posted: Thu Mar 21, 2019 4:41 pm
by cdienger
I'd like to take a closer look at the plugin that is generating this output. What is the plugin and command you are running to get it?

Re: Memory usage showing wrong Alert

Posted: Fri Mar 22, 2019 2:55 am
by manimurugesan
Hello

Please find the below command definition we are using check_nrpe and command is "custom_check_mem"

Re: Memory usage showing wrong Alert

Posted: Fri Mar 22, 2019 2:39 pm
by cdienger
The check uses the following to get values:

Code: Select all

$svmon -G | grep "^memory" | cut -f 2 -d " "
$svmon -G | grep "^memory" | cut -f 3 -d " "
$svmon -G | grep "^memory" | cut -f 4 -d " "
$svmon -G | grep "^memory" | cut -f 5 -d " "
It then divides each value by 256 to get the total, inuse, free, and pinned values. Try running these directly on the AIX machine to get the output and then run the check again and compare it to the output.

Note that it appears to drop everything beyond the decimal place and 6306/57344 = 10.9997% and 5734/57344 = 9.999%

Re: Memory usage showing wrong Alert

Posted: Wed Mar 27, 2019 4:00 am
by manimurugesan
Hello,

The issue has been resolved.

Server end free space was 10.3% but in nagios threshold limit was 10% so server end 11% memory should be free because it is not taking the decimal values like 10.1,10.2 .

Re: Memory usage showing wrong Alert

Posted: Wed Mar 27, 2019 8:09 am
by scottwilkerson
manimurugesan wrote:Hello,

The issue has been resolved.

Server end free space was 10.3% but in nagios threshold limit was 10% so server end 11% memory should be free because it is not taking the decimal values like 10.1,10.2 .
Great!

Locking thread