hi HI im using DISK * 97 99 ANd still dont works. Now it works to you? where had you change the df command? at the server? what file? Thanks u very much!

2008/10/8 Engineering <forums at triadbiz.rr.com>:

I believe I have found the issue. I changed on my client scripts the "-l" to a "-h" to make the readings human readable. Well in doing that it changed the Column names across the top of the command:

[default hobbit command] df -Pl Filesystem 1024-blocks Used Available Capacity Mounted on /dev/mapper/VolGroup00-LogVol00 32344948 23893632 6781788 78% / /dev/sda1 99134 17731 76284 19% /boot tmpfs 1684900 0 1684900 0% /dev/shm

[my modification for the command] df -Ph Filesystem Size Used Avail Use% Mounted on /dev/mapper/VolGroup00-LogVol00 31G 23G 6.5G 78% / /dev/sda1 97M 18M 75M 19% /boot tmpfs 1.7G 0 1.7G 0% /dev/shm gsonc-spg-snap:/OPS1 1.9T 276G 1.6T 15% /data

So what I did is changed a line in "linux.c" and recompiled hobbit. Now it is alarming like it should. Might need to look into a different way for parsing the DISK info.

Thanks,

-Josh

-----Original Message----- From: jose erquicia [mailto:mosimodo at gmail.com] Sent: Wednesday, October 08, 2008 10:37 AM To: hobbit at hswn.dk Subject: Re: [hobbit] monitoring DISK with hobbit

hi ! mi problem is the same i´m doing tests but can´t find anything. If i found the solution I will tell you. regards!

2008/10/8 Engineering <forums at triadbiz.rr.com>:

...

I am having issues monitoring the disk usage via hobbit. It is not alerting out when the usage goes over limits. Here is what I have in my hobbit-clients.cfg

HOST=%gsonc-spg-bms-1

    PROC    BmsUI "TEXT=BMS Web UI"

    PROC    CORBAnameservice "TEXT=Name Service"

    PROC    CORBAnotificationservice "TEXT=Notification Service"

    PROC    SettingsServer "TEXT=Setting Server"

HOST=%gsonc-spg-bms-2

    PROC    CBITrans "TEXT=CBI Transactions"

    PROC    Customer

    PROC    ServiceGateway "TEXT=Service Gateway"

    PROC    SessionGateway "TEXT=Session Gateway"

HOST=%gsonc-spg-bms-3

    PROC    Metadata

    PROC    Package

    PROC    Product

    PROC    Terminal

HOST=%gsonc-spg-bms-4

    PROC    Asset

    PROC    Provider

    PROC    Purchase

HOST=%gsonc-spg-ops-1

    LOAD    12.0 17.0

    PROC    snmpd

    PROC    snmptt

    PROC    sec.pl

    PROC    snmptrapd

    PROC    mysqld

HOST=%gsonc-spg-ops-2

    PROC    radiusd

    PROC    mysqld

HOST=%gsonc-spg-ops-3

    PROC    syslog-ng

    PROC    mysqld

HOST=%gsonc-spg-ops-4

    PROC    radiusd

    PROC    mysqld

HOST=%gsonc-spg-omea-1

    PROC    cron

    LOAD    7.0 10.0

HOST=%gsonc-spg-sectimus.*

    PROC    sectamus

HOST=%gsonc-spg-sunbox

    LOAD    11.5 12.5

    MEMSWAP 85 95

HOST=%gsonc-spg-dncs

    FILE    /export/home/dncs/mystro/bfs/Compass_OOB/rta.grp.z

SIZE>1200 TRACK=rta

    LOAD    10.0 15.0

    DISK    / 95 98

    DISK    /cdrom IGNORE

HOST=%gsonc-spg-mas*

    FILE /usr/local/mystro/conf/rtareports_headendinfo.properties

MTIME=1128514608 TRACK=rta-reports

    MEMPHYS 100 101

    MEMSWAP 75 90

    MEMACT  97 98

PAGE=%.*/vod

    LOAD    17.0 22.0

    DISK    / 50 95

DEFAULT

    # These are the built-in defaults.

    UP      1h

    LOAD    17.0 22.0

    DISK    / 80 95

    MEMPHYS 100 101

    MEMSWAP 75 90

    MEMACT  90 97

I have tried the HOST and PAGE values and still can't get anything to work. The weird thing is LOAD, MEMORY, etc work just not DISK. Is there a known bug with hobbit and DISK or am I doing something wrong.

Thanks,

-Josh

To unsubscribe from the hobbit list, send an e-mail to hobbit-unsubscribe at hswn.dk

To unsubscribe from the hobbit list, send an e-mail to hobbit-unsubscribe at hswn.dk

In <200810090944.29145.bgmilne at staff.telkomsa.net> Buchan Milne <bgmilne at staff.telkomsa.net> writes:

On Wednesday 08 October 2008 19:05:15 Engineering wrote:

...

I believe I have found the issue. I changed on my client scripts the "-l" to a "-h" to make the readings human readable.

I don't think Henrik intended this, and it's more than a one-line patch to make it possible to use disk utilisation displayed in the human-readable form

Maybe we could have the clients report both ways - using the "-P" option for feeding data into the graphs, checking utilisation against the client config and so on - and have a different command whose output is displayed on the disk status ?

It wouldn't be difficult to implement, but my worry is that there might be a situation where the two sets of data are slightly different (e.g. rounding errors, or a rapidly growing file between the two measurements) and that will be confusing.

Henrik