Host processes monitoring-graphs LLD

[ehermouet]

with serveractive=ip it's work tks

[MrSol]

roll-out on new agent

Tried to roll-out scripts on new agent and had same issues with initial creation of zabbix_win_process.saved file.
Command zabbix_win_process.vbs read svchost.exe gave empty output. Agent was restarted several times without success.
I was suspecting WMI and tried to reset the service.

To manually reset the WMI counters:
1. Click Start, click Run, type cmd, and then click OK.
2. At the command prompt, type winmgmt /resyncperf, and then press
ENTER.
3. At the command prompt, type wmiadap.exe /f, and then press ENTER.
4. Type exit, and then press ENTER to close the command prompt.
5. Restart the Windows Management Instrumentation service.

Then zabbix_win_process.vbs commit svchost.exe generated zabbix_win_process.saved file with single process inside. I restarted an agent afterwards and waited the next poll. Finally got top 9 processes (by memory size) in the file.

Hope that'll help someone with similar issues.

[mmarkwitzz]

here my log on agent

6112:20120731:161028.011 Active check [eventlog.query[application,(source=Application\sError;id=1000;typ e=Error) | (source=Application\sHang;id=1002;type=Error)]] is not supported. Disabled.
6112:20120731:161028.011 Active check [eventlog.query[oop,oop,,]] is not supported. Disabled.
6112:20120731:161028.011 Active check [eventlog.query[system,(source=Disk;id=11|7;type=Error) | (source=Disk;id=52|51;type=Warning) | (source=Ntfs;id=137|55;type=Error) | (source=Ntfs;id=136|57;type=Warning) | (source=Virtual\sDisk\sService;id=9|8|6|1;type=Err or)]] is not supported. Disabled.
6112:20120731:161028.011 Active check [eventlog.query[system,(source=LMDS;id=3014;type=Error) | (source=LMD;id=2;type=Error) | (source=atapi;id=9;type=Error) | (source=atapi;id=26;type=Warning) | (source=iaStor;id=9;type=Error) | (source=dmio;id=36|5|4;type=Warning)]] is not supported. Disabled.
6112:20120731:161031.074 Active check [system.discovery[disks,]] is not supported. Disabled.
6112:20120731:161034.090 Active check [system.discovery[disks]] is not supported. Disabled.
6112:20120731:161037.105 Active check [system.discovery[processes]] is not supported. Disabled.
6112:20120731:161040.480 Active check [system.uptime[]] is not supported. Disabled.

Edit your config file and enable UnsafeUserParameters=1 (go to documentation, I'm not 100% this is the correct parameter name)

Edit: At first sight it would seem that either the UserParameters are not defined in you .conf file or that the paths to the scripts are wrong.
But the log also states that system.uptime[] is unsupported which is really odd as it's a check built into the agent itself.
Check the UserParameters and make sure they are corrent and the paths to the scripts are correct, and if you still have the issue, raise the DebugLevel to 3 or 4 and get more details on _why_ the checks are not supported.

[tchjts1]

So I have your mods working fine, with one exception...

I create a screen of graphs, making each graph a dynamic item. The screen works for a period of time, then *poof*, the screen is blank and all graphs are gone.

Is it not possible to have a screen of these graphs? Is it because the pl scripts are creating new graphs every hour? Any way around this?

I am not seeing the issue on other dynamic screens of regular Zabbix item graphs.

[MrSol]

I create a screen of graphs, making each graph a dynamic item. The screen works for a period of time, then *poof*, the screen is blank and all graphs are gone.

Same here. What is the workaround?

[mmarkwitzz]

The script deletes the previous graphs (if any exists) and created new ones. And the screen you created has references to the IDs of the deleted graphs.

I cannot help you with a workaround, I have changed jobs and I haven't worked with zabbix for 4 months.

A workaround for this would be to also script the creation of the screen, or to script the change of that particular graph in the screen.

[tchjts1]

I cannot help you with a workaround, I have changed jobs and I haven't worked with zabbix for 4 months.

Congrats on the new job! (Hope it was for the better!)
Your contributions to the Zabbix community will be missed.

[tchjts1]

Is anyone else observing negative amounts on the processor time stack graph? Any way to correct this?

[ctreppish]

So I have your mods working fine, with one exception...

I create a screen of graphs, making each graph a dynamic item. The screen works for a period of time, then *poof*, the screen is blank and all graphs are gone.

Is it not possible to have a screen of these graphs? Is it because the pl scripts are creating new graphs every hour? Any way around this?

I am not seeing the issue on other dynamic screens of regular Zabbix item graphs.

What did you do to get past the malformed JSON string error. I am having the same problem. I am running this off the Zabbix appliance running 2.0.2

Thanks

[tchjts1]

What did you do to get past the malformed JSON string error. I am having the same problem. I am running this off the Zabbix appliance running 2.0.2

Thanks

More than likely the issue is with your $url parameter inside of each perl script. In my case, we use port 10052 to the frontend, so my url path needed to include that.

it looked like $url = "http://123.123.123.123:10052/api_jsonrpc.php"; ### intenal zabbix url

If you are using 127.0.0.1, try using the actual IP address of your server

[nickp]

Hi

When i try to execute the perl script on the zabbix server i get this:

]# Can't locate JSON.pm in @INC (@INC contains: /usr/local/lib64/perl5 /usr/local/share/perl5 /usr/lib64/perl5/vendor_perl /usr/share/perl5/vendor_perl /usr/lib64/perl5 /usr/share/perl5 .) at win_process_all_bytessec_stack.pl line 2.
BEGIN failed--compilation aborted at win_process_all_bytessec_stack.pl line 2.

anyone have an idea? CentOS 6.3

[tchjts1]

Sounds like you are missing the json module.

It wasn't listed in my repositories, so I downloaded the RPM and installed that. This is the package name for RHEL6:

perl-JSON.noarch 2.50-1.el6.rfx

[nickp]

Thx, yum installed it, and no more errors.

However no graphs are generated, after running the pl scripts.

[root@secit zabbix]# ./win_process_all_threads_stack.pl

HOSTGROUP: Database Servers (11)
HOSTGROUP: Desktops (12)
HOSTGROUP: Domain Controllers (10)
HOSTGROUP: Fujitsu Servers (7)
HOSTGROUP: Linux servers (2)
HOSTGROUP: Mail servers (6)
HOSTGROUP: Storage Servers (16)
HOSTGROUP: Switches (14)
HOSTGROUP: Terminalservers (13)
HOSTGROUP: Virtual Hosts (17)
HOSTGROUP: VOIP (15)
HOSTGROUP: Zabbix servers (4)
HOSTGROUP: ZTM Recomatics (9)
HOST: ZTM-SRV-AUTOMAT (10124)
Graph created: WIN Process "ALL" threads stack
Is all i get...

[tchjts1]

I think that is the right output. You linked the template to your Windows hosts, right?

Then go to Monitoring --> Graphs and select the host. You should see the graphs listed there.

[ctreppish]

More than likely the issue is with your $url parameter inside of each perl script. In my case, we use port 10052 to the frontend, so my url path needed to include that.

it looked like $url = "http://123.123.123.123:10052/api_jsonrpc.php"; ### intenal zabbix url

If you are using 127.0.0.1, try using the actual IP address of your server

Are the user name and password what we use to access the console?

For the url sting I have tried the following with no luck

"http: // 10.178.7.102 /zabbix/api_jsonrpc.php";
"http: // 10.178.7.102 /api_jsonrpc.php";
"http: // 10.178.7.102:10051 /api_jsonrpc.php";

"http: // 127.0.0.1 /zabbix/api_jsonrpc.php";
"http: // 127.0.0.1 /api_jsonrpc.php";

Any help would be greatly appreciated.

This is how my perl script looks:


$user = "Admin";

$password = "zabbix";

$template = "WIN Processes";

$header = "Content-Type:application/json";

$url = "http:// 10.178.7.102 /api_jsonrpc.php";