zabbix-users — For general discussions

Re: [Zabbix-users] Returning an error message from an externalscript

[Brian C. <b.c...@po...>]

On 08/02/2017 11:48, Brian Candler wrote:
> I note that Zabbix has a concept of "unsupported" value, but it 
> doesn't say how an external script could return this value (empty 
> string? Non-zero exit code?) nor how an error string could be 
> associated with an unsupported value and included in an alert action. 

Ah, it looks like maybe I am looking for this new feature in 
(not-yet-released) 3.4.0:

https://www.zabbix.com/documentation/3.4/manual/appendix/command_execution#exit_code_checking

"Any exit code that is different from 0 is considered as execution failure."

Technically, execution failure (of the script itself) is not the same as 
"bad response to test" (from the remote host); but I don't think that 
matters here.

Re: [Zabbix-users] Zabbix Agent on SoftLayer Virtual machine

[Michael N. <mic...@gm...>]

It seems the template are using
vm.memory.size[total]
and
vm.memory.size[available]


On Wed, Feb 8, 2017 at 2:59 PM, Michael Nielsen <mic...@gm...>
wrote:

> Hmm, simply used the "Template Linux OS". That has some memory data in it.
>
> On Wed, Feb 8, 2017 at 11:08 AM, Jeon Woosung <jeo...@gm...>
> wrote:
>
>> Have you used "pused" instead of "pavailable"?
>>
>> 2017. 2. 8. 오후 5:13에 "Michael Nielsen" <mic...@gm...>님이 작성:
>>
>>> Running free -m gives me:
>>>
>>>              total       used       free     shared    buffers     cached
>>>
>>> Mem:          3958       2518       1440          8        147       2179
>>>
>>> It is the used 2.5Gb and free 1.5Gb I would like Zabbix to show.
>>>
>>> Zabbix gives me:
>>>
>>> Available memory 2017-02-08 08:06:30 3.55 GB -4.71 MB Graph
>>> Free swap space 2017-02-08 08:06:12 2 GB Graph
>>> Free swap space in % 2017-02-08 08:06:23 100 % Graph
>>> Total memory 2017-02-08 07:08:16 3.87 GB Graph
>>> Total swap space 2017-02-08 07:08:17 2 GB Graph
>>>
>>> So not the same as free -m
>>>
>>>
>>> On Tue, Feb 7, 2017 at 8:28 AM, Michael Nielsen <mic...@gm...>
>>> wrote:
>>>
>>>> Thank you. :)
>>>> I'll play around with zabbix.
>>>>
>>>> On Monday, February 6, 2017, Guus Snijders <gsn...@gm...> wrote:
>>>>
>>>>> Op 6 feb. 2017 22:34 schreef "Michael Nielsen" <mic...@gm...>:
>>>>>
>>>>> I see.
>>>>> Is there a way to see memory used (as regular understood?) or free
>>>>> memory which applications can use?
>>>>>
>>>>>
>>>>> Sure, start with free -m, study it's output (can be confusing!) so
>>>>> you'll get some feeling for the different fields.
>>>>>
>>>>> I have one VM with 4GB allowed.
>>>>> Running top in the command line tells me 2,5 is used and there's 1,5
>>>>> available.
>>>>> This is what I want have shown in zabbix  graph.
>>>>>
>>>>>
>>>>> Can't tell from the top of my head, sry.
>>>>>
>>>>> Luckily, the agent has good documentation, so find out which info
>>>>> (metrics) you want and then check the the wiki for supported items. Worst
>>>>> case, it's simple to write a one-liner that gives the info, but the agent
>>>>> probably already builtin ;).
>>>>>
>>>>> It could very well be something like mem.avail, but that's just a wild
>>>>> guess.
>>>>>
>>>>> Note, this is one of the best ways to get to know Zabbix; starting
>>>>> with standard templates and then rolling your own.
>>>>>
>>>>>
>>>>> Just my € 0,02
>>>>>
>>>>> Mvg, Guus Snijders
>>>>>
>>>>>
>>>>>
>>>
>>> ------------------------------------------------------------
>>> ------------------
>>> Check out the vibrant tech community on one of the world's most
>>> engaging tech sites, SlashDot.org! http://sdm.link/slashdot
>>> _______________________________________________
>>> Zabbix-users mailing list
>>> Zab...@li...
>>> https://lists.sourceforge.net/lists/listinfo/zabbix-users
>>>
>>>
>

Re: [Zabbix-users] Zabbix Agent on SoftLayer Virtual machine

[Michael N. <mic...@gm...>]

Hmm, simply used the "Template Linux OS". That has some memory data in it.

On Wed, Feb 8, 2017 at 11:08 AM, Jeon Woosung <jeo...@gm...> wrote:

> Have you used "pused" instead of "pavailable"?
>
> 2017. 2. 8. 오후 5:13에 "Michael Nielsen" <mic...@gm...>님이 작성:
>
>> Running free -m gives me:
>>
>>              total       used       free     shared    buffers     cached
>>
>> Mem:          3958       2518       1440          8        147       2179
>>
>> It is the used 2.5Gb and free 1.5Gb I would like Zabbix to show.
>>
>> Zabbix gives me:
>>
>> Available memory 2017-02-08 08:06:30 3.55 GB -4.71 MB Graph
>> Free swap space 2017-02-08 08:06:12 2 GB Graph
>> Free swap space in % 2017-02-08 08:06:23 100 % Graph
>> Total memory 2017-02-08 07:08:16 3.87 GB Graph
>> Total swap space 2017-02-08 07:08:17 2 GB Graph
>>
>> So not the same as free -m
>>
>>
>> On Tue, Feb 7, 2017 at 8:28 AM, Michael Nielsen <mic...@gm...>
>> wrote:
>>
>>> Thank you. :)
>>> I'll play around with zabbix.
>>>
>>> On Monday, February 6, 2017, Guus Snijders <gsn...@gm...> wrote:
>>>
>>>> Op 6 feb. 2017 22:34 schreef "Michael Nielsen" <mic...@gm...>:
>>>>
>>>> I see.
>>>> Is there a way to see memory used (as regular understood?) or free
>>>> memory which applications can use?
>>>>
>>>>
>>>> Sure, start with free -m, study it's output (can be confusing!) so
>>>> you'll get some feeling for the different fields.
>>>>
>>>> I have one VM with 4GB allowed.
>>>> Running top in the command line tells me 2,5 is used and there's 1,5
>>>> available.
>>>> This is what I want have shown in zabbix  graph.
>>>>
>>>>
>>>> Can't tell from the top of my head, sry.
>>>>
>>>> Luckily, the agent has good documentation, so find out which info
>>>> (metrics) you want and then check the the wiki for supported items. Worst
>>>> case, it's simple to write a one-liner that gives the info, but the agent
>>>> probably already builtin ;).
>>>>
>>>> It could very well be something like mem.avail, but that's just a wild
>>>> guess.
>>>>
>>>> Note, this is one of the best ways to get to know Zabbix; starting with
>>>> standard templates and then rolling your own.
>>>>
>>>>
>>>> Just my € 0,02
>>>>
>>>> Mvg, Guus Snijders
>>>>
>>>>
>>>>
>>
>> ------------------------------------------------------------
>> ------------------
>> Check out the vibrant tech community on one of the world's most
>> engaging tech sites, SlashDot.org! http://sdm.link/slashdot
>> _______________________________________________
>> Zabbix-users mailing list
>> Zab...@li...
>> https://lists.sourceforge.net/lists/listinfo/zabbix-users
>>
>>

Re: [Zabbix-users] Avoiding long keys with system.run

[Guus S. <gsn...@gm...>]

Op 8 feb. 2017 13:03 schreef "Brian Candler" <b.c...@po...>:

Hello,

I have a question about sensible deployment of Zabbix.

I've found that a simple way to check certain conditions is to create an
item which runs a remote command on the host. For example, if I want a
check that a host has password authentication disabled, I can create a
template with an item like this:

Type: Zabbix Agent
Key: system.run["grep '^PasswordAuthentication no' /etc/ssh/sshd_config |
wc -l"]

[Snip trigger, long key]

I fell into this "trap" too ;).
In my case it turned out that Zabbix was the wrong tool for the job.

Is this something you want to check regularly? If yes: use a custom key in
the agent config, active/passive according to taste and execution time.

If it's something you want to know once (or every now and than),
ssh/puppet/salt/whatever may very well be the better choice.

For a once-off, a ssh loop in bash is probably a good option. Zabbix is
very good in repeating actions every $interval.
One-time actions are possible, but usually end up cluttering the db...


Mvg, Guus Snijders

Re: [Zabbix-users] Avoiding long keys with system.run

[Todd B. <tb...@gm...>]

In regards to this problem:
*If I decided to change the implementation of the PasswordAuthentication
check, I'd have to change any related triggers too (although as they're
likely in the same template, that's not too bad).*

Not necessarily.  When you update the item, it'll update the trigger.

The other option is to create a custom key on the agent, say call it
custom.ssh.passauth, and have it run the command, with a directive like:

UserParameter=custom.ssh.passauth,grep '^PasswordAuthentication no'
/etc/ssh/sshd_config | wc -l


in your zabbix_agentd.conf file.  Then you can just make your key in the
Zabbix UI *custom.ssh.passauth*.  But then you're maintaining it on all of
your agents.  Presumable via puppet/chef/whatever.  So it'll still need to
be maintained somewhere.  Ideally though, you define it once on a template
in the item, define the trigger, and any item updates automatically update
the trigger.  Then like you said, that's in a template, so you're only
doing it once.

On Wed, Feb 8, 2017 at 6:32 AM, Brian Candler <b.c...@po...> wrote:

> Hello,
>
> I have a question about sensible deployment of Zabbix.
>
> I've found that a simple way to check certain conditions is to create an
> item which runs a remote command on the host. For example, if I want a
> check that a host has password authentication disabled, I can create a
> template with an item like this:
>
> Type: Zabbix Agent
> Key: system.run["grep '^PasswordAuthentication no' /etc/ssh/sshd_config |
> wc -l"]
>
> This is simple to implement, but the problem is that the *key* used to
> refer to that data is that very long string.  If I want to write a trigger
> for it, the condition looks something like this:
>
> {Linux Security Compliance Rules:system.run["grep '^PasswordAuthentication
> no' /etc/ssh/sshd_config | wc -l"].last()}=0
>
> This doesn't feel right. I'm duplicating the long string which runs the
> command. If I decided to change the implementation of the
> PasswordAuthentication check, I'd have to change any related triggers too
> (although as they're likely in the same template, that's not too bad).
>
> Have I missed a trick here? Is there a way to define a command to run on
> the remote system, but give it a shorter key name?
> Thanks,
>
> Brian.
>
>
> ------------------------------------------------------------
> ------------------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, SlashDot.org! http://sdm.link/slashdot
> _______________________________________________
> Zabbix-users mailing list
> Zab...@li...
> https://lists.sourceforge.net/lists/listinfo/zabbix-users
>
>

[Zabbix-users] Returning an error message from an externalscript

[Brian C. <b.c...@po...>]

Suppose I write a custom external script 
/usr/lib/zabbix/externalscripts/foo.sh and create an item for it like this:

Type: External Script
Key: foo.sh[{HOST.IP}]

The result from the script is normally numeric (let's say a response 
time), and I might want to graph this or alert on the value exceeding a 
threshold.

However it can also return an error status as a text string. There might 
be several different error strings (e.g. "Connection failed", "DNS 
resolution failed", "Unexpected response: 501"). I want to send out 
alerts in those conditions too, and I want the alerts to contain the 
message returned by the script.

What's the best way to implement this, in terms of what the script 
outputs, and what the corresponding item type is in Zabbix?

The options I can see:

(1) I could configure it with "Type of information: Numeric (float)", 
but then I don't know what happens when the script outputs an error message

(2) I could configure it as "Type of information: Text". Then I can 
write actions which match numeric-only or non-numeric values, but that 
requires regexp checks and it seems pretty messy; I'm also not sure I 
can compare a text value to a numeric threshold.

Presumably this is a very common situation, so what's the recommended 
way to deal with it?

I've read the documentation here:
https://www.zabbix.com/documentation/3.2/manual/config/items/itemtypes/external

I note that Zabbix has a concept of "unsupported" value, but it doesn't 
say how an external script could return this value (empty string? 
Non-zero exit code?) nor how an error string could be associated with an 
unsupported value and included in an alert action.

Thanks,

Brian.

[Zabbix-users] Avoiding long keys with system.run

[Brian C. <b.c...@po...>]

Hello,

I have a question about sensible deployment of Zabbix.

I've found that a simple way to check certain conditions is to create an 
item which runs a remote command on the host. For example, if I want a 
check that a host has password authentication disabled, I can create a 
template with an item like this:

Type: Zabbix Agent
Key: system.run["grep '^PasswordAuthentication no' /etc/ssh/sshd_config 
| wc -l"]

This is simple to implement, but the problem is that the *key* used to 
refer to that data is that very long string.  If I want to write a 
trigger for it, the condition looks something like this:

{Linux Security Compliance Rules:system.run["grep 
'^PasswordAuthentication no' /etc/ssh/sshd_config | wc -l"].last()}=0

This doesn't feel right. I'm duplicating the long string which runs the 
command. If I decided to change the implementation of the 
PasswordAuthentication check, I'd have to change any related triggers 
too (although as they're likely in the same template, that's not too bad).

Have I missed a trick here? Is there a way to define a command to run on 
the remote system, but give it a shorter key name?

Thanks,

Brian.

Re: [Zabbix-users] Zabbix Agent on SoftLayer Virtual machine

[Jeon W. <jeo...@gm...>]

Have you used "pused" instead of "pavailable"?

2017. 2. 8. 오후 5:13에 "Michael Nielsen" <mic...@gm...>님이 작성:

> Running free -m gives me:
>
>              total       used       free     shared    buffers     cached
>
> Mem:          3958       2518       1440          8        147       2179
>
> It is the used 2.5Gb and free 1.5Gb I would like Zabbix to show.
>
> Zabbix gives me:
>
> Available memory 2017-02-08 08:06:30 3.55 GB -4.71 MB Graph
> Free swap space 2017-02-08 08:06:12 2 GB Graph
> Free swap space in % 2017-02-08 08:06:23 100 % Graph
> Total memory 2017-02-08 07:08:16 3.87 GB Graph
> Total swap space 2017-02-08 07:08:17 2 GB Graph
>
> So not the same as free -m
>
>
> On Tue, Feb 7, 2017 at 8:28 AM, Michael Nielsen <mic...@gm...>
> wrote:
>
>> Thank you. :)
>> I'll play around with zabbix.
>>
>> On Monday, February 6, 2017, Guus Snijders <gsn...@gm...> wrote:
>>
>>> Op 6 feb. 2017 22:34 schreef "Michael Nielsen" <mic...@gm...>:
>>>
>>> I see.
>>> Is there a way to see memory used (as regular understood?) or free
>>> memory which applications can use?
>>>
>>>
>>> Sure, start with free -m, study it's output (can be confusing!) so
>>> you'll get some feeling for the different fields.
>>>
>>> I have one VM with 4GB allowed.
>>> Running top in the command line tells me 2,5 is used and there's 1,5
>>> available.
>>> This is what I want have shown in zabbix  graph.
>>>
>>>
>>> Can't tell from the top of my head, sry.
>>>
>>> Luckily, the agent has good documentation, so find out which info
>>> (metrics) you want and then check the the wiki for supported items. Worst
>>> case, it's simple to write a one-liner that gives the info, but the agent
>>> probably already builtin ;).
>>>
>>> It could very well be something like mem.avail, but that's just a wild
>>> guess.
>>>
>>> Note, this is one of the best ways to get to know Zabbix; starting with
>>> standard templates and then rolling your own.
>>>
>>>
>>> Just my € 0,02
>>>
>>> Mvg, Guus Snijders
>>>
>>>
>>>
>
> ------------------------------------------------------------
> ------------------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, SlashDot.org! http://sdm.link/slashdot
> _______________________________________________
> Zabbix-users mailing list
> Zab...@li...
> https://lists.sourceforge.net/lists/listinfo/zabbix-users
>
>

Re: [Zabbix-users] Zabbix Agent on SoftLayer Virtual machine

[Michael N. <mic...@gm...>]

Running free -m gives me:

             total       used       free     shared    buffers     cached

Mem:          3958       2518       1440          8        147       2179

It is the used 2.5Gb and free 1.5Gb I would like Zabbix to show.

Zabbix gives me:

Available memory 2017-02-08 08:06:30 3.55 GB -4.71 MB Graph
Free swap space 2017-02-08 08:06:12 2 GB Graph
Free swap space in % 2017-02-08 08:06:23 100 % Graph
Total memory 2017-02-08 07:08:16 3.87 GB Graph
Total swap space 2017-02-08 07:08:17 2 GB Graph

So not the same as free -m


On Tue, Feb 7, 2017 at 8:28 AM, Michael Nielsen <mic...@gm...>
wrote:

> Thank you. :)
> I'll play around with zabbix.
>
> On Monday, February 6, 2017, Guus Snijders <gsn...@gm...> wrote:
>
>> Op 6 feb. 2017 22:34 schreef "Michael Nielsen" <mic...@gm...>:
>>
>> I see.
>> Is there a way to see memory used (as regular understood?) or free memory
>> which applications can use?
>>
>>
>> Sure, start with free -m, study it's output (can be confusing!) so you'll
>> get some feeling for the different fields.
>>
>> I have one VM with 4GB allowed.
>> Running top in the command line tells me 2,5 is used and there's 1,5
>> available.
>> This is what I want have shown in zabbix  graph.
>>
>>
>> Can't tell from the top of my head, sry.
>>
>> Luckily, the agent has good documentation, so find out which info
>> (metrics) you want and then check the the wiki for supported items. Worst
>> case, it's simple to write a one-liner that gives the info, but the agent
>> probably already builtin ;).
>>
>> It could very well be something like mem.avail, but that's just a wild
>> guess.
>>
>> Note, this is one of the best ways to get to know Zabbix; starting with
>> standard templates and then rolling your own.
>>
>>
>> Just my € 0,02
>>
>> Mvg, Guus Snijders
>>
>>
>>

Re: [Zabbix-users] Zabbix Agent on SoftLayer Virtual machine

[Michael N. <mic...@gm...>]

Thank you. :)
I'll play around with zabbix.

On Monday, February 6, 2017, Guus Snijders <gsn...@gm...> wrote:

> Op 6 feb. 2017 22:34 schreef "Michael Nielsen" <mic...@gm...
> <javascript:_e(%7B%7D,'cvml','mic...@gm...');>>:
>
> I see.
> Is there a way to see memory used (as regular understood?) or free memory
> which applications can use?
>
>
> Sure, start with free -m, study it's output (can be confusing!) so you'll
> get some feeling for the different fields.
>
> I have one VM with 4GB allowed.
> Running top in the command line tells me 2,5 is used and there's 1,5
> available.
> This is what I want have shown in zabbix  graph.
>
>
> Can't tell from the top of my head, sry.
>
> Luckily, the agent has good documentation, so find out which info
> (metrics) you want and then check the the wiki for supported items. Worst
> case, it's simple to write a one-liner that gives the info, but the agent
> probably already builtin ;).
>
> It could very well be something like mem.avail, but that's just a wild
> guess.
>
> Note, this is one of the best ways to get to know Zabbix; starting with
> standard templates and then rolling your own.
>
>
> Just my € 0,02
>
> Mvg, Guus Snijders
>
>
>

Re: [Zabbix-users] Zabbix Agent on SoftLayer Virtual machine

[Guus S. <gsn...@gm...>]

Op 6 feb. 2017 22:34 schreef "Michael Nielsen" <mic...@gm...>:

I see.
Is there a way to see memory used (as regular understood?) or free memory
which applications can use?


Sure, start with free -m, study it's output (can be confusing!) so you'll
get some feeling for the different fields.

I have one VM with 4GB allowed.
Running top in the command line tells me 2,5 is used and there's 1,5
available.
This is what I want have shown in zabbix  graph.


Can't tell from the top of my head, sry.

Luckily, the agent has good documentation, so find out which info (metrics)
you want and then check the the wiki for supported items. Worst case, it's
simple to write a one-liner that gives the info, but the agent probably
already builtin ;).

It could very well be something like mem.avail, but that's just a wild
guess.

Note, this is one of the best ways to get to know Zabbix; starting with
standard templates and then rolling your own.


Just my € 0,02

Mvg, Guus Snijders

Re: [Zabbix-users] Zabbix Agent on SoftLayer Virtual machine

[Michael N. <mic...@gm...>]

I see.
Is there a way to see memory used (as regular understood?) or free memory
which applications can use?

I have one VM with 4GB allowed.
Running top in the command line tells me 2,5 is used and there's 1,5
available.
This is what I want have shown in zabbix  graph.

On Monday, February 6, 2017, Guus Snijders <gsn...@gm...> wrote:

> Op 6 feb. 2017 13:23 schreef "Michael Nielsen" <mic...@gm...
> <javascript:_e(%7B%7D,'cvml','mic...@gm...');>>:
>
> I'm running a few Debian virtual machines on IBM SoftLayer.
> And things seems to work, but for the memory usage I'm getting almost 90%
> free on all machine, and I know a few of them uses over 60 %.
>
>
> Metrics can be confusing.
> Try running:
> free -m
> In one of those vm's and compare the values against zabbix & SoftLayer.
> What you'll probably see is that the hypervisor reports all allocated
> memory as "used" (including buffers/cache), whereas zabbix shows how much
> memory can be available for applications.
>
> Funny detail: both are correct. They just show different metrics with
> different meanings.
>
>
> Mvg, Guus Snijders
>
>

Re: [Zabbix-users] Zabbix Agent on SoftLayer Virtual machine

[Guus S. <gsn...@gm...>]

Op 6 feb. 2017 13:23 schreef "Michael Nielsen" <mic...@gm...>:

I'm running a few Debian virtual machines on IBM SoftLayer.
And things seems to work, but for the memory usage I'm getting almost 90%
free on all machine, and I know a few of them uses over 60 %.


Metrics can be confusing.
Try running:
free -m
In one of those vm's and compare the values against zabbix & SoftLayer.
What you'll probably see is that the hypervisor reports all allocated
memory as "used" (including buffers/cache), whereas zabbix shows how much
memory can be available for applications.

Funny detail: both are correct. They just show different metrics with
different meanings.


Mvg, Guus Snijders

[Zabbix-users] Zabbix Agent on SoftLayer Virtual machine

[Michael N. <mic...@gm...>]

I'm running a few Debian virtual machines on IBM SoftLayer.
And things seems to work, but for the memory usage I'm getting almost 90%
free on all machine, and I know a few of them uses over 60 %.

I tried to google around, and it seems there can be something with virtual
machines and Zabbix (and in general monitoring tools).

I'm using the "Template OS Linux".

I'm a newbie to Zabbix.

What should I do?

Re: [Zabbix-users] Debug LLD with trapper item prototype

[Michael L. <mcs...@gm...>]

Here is little quote from Your letter:

# *Item prototypes 2  without zabbix result*
  Key:  lvs_service[inbps,{$SERVICE_HOST},{$SERVICE_PORT}]
  Type: Zabbix trapper

May be, it would be better to replace $ in item key with #?

2017-01-24 4:38 GMT+03:00 linbo liao <ll...@gm...>:

> Hi,
>
> In my test environment I try to setup lvs monitor with zabbix LLD.
>
> Zabbix 3.2.0
> OS: Ubuntu 14.04.1 LTS
>
> *# Discovery rule*
> 1. key: lvs.service
> 2. Type: Zabbix agent
> 3. Two Filters with empty regular expression
>     {#SERVICE_HOST}
>     {#SERVICE_PORT}
>
> # cat /etc/zabbix/zabbix_agentd.conf.d/userparameter_lvs.conf
>> UserParameter=lvs.service, python /etc/zabbix/script/lvs_
>> monitor/dump_service.py
>>
>
> # python /etc/zabbix/script/lvs_monitor/dump_service.py
>> {"data": [{"{#SERVICE_HOST}": "10.211.55.12", "{#SERVICE_PORT}": "7777"},
>> {"{#SERVICE_HOST}": "10.211.55.11", "{#SERVICE_PORT}": "7778"}]}
>
>
> *# Item prototypes 1 works fine*
>   Key: net.tcp.service[tcp,{#SERVICE_HOST},{#SERVICE_PORT}]
>   Type: Zabbix agent
>
> # *Item prototypes 2  without zabbix result*
>   Key:  lvs_service[inbps,{$SERVICE_HOST},{$SERVICE_PORT}]
>   Type: Zabbix trapper
>
> A crontab use zabbix_sender to send data to zabbix_server
>
> # cat /tmp/lvs_service_stat
>> precise64 lvs_service[inpps,10.211.55.12,7777] 1485221745 0
>> precise64 lvs_service[inpps,10.211.55.11,7778] 1485221745 0
>>
>
>
>
>
> *The Zabbix Server log*
>>
>> trapper got '{
>>     "request":"sender data",
>>     "data":[
>>         {
>>             "host":"precise64",
>>             "key":"lvs_service[inpps,10.211.55.12,7777]",
>>             "value":"0",
>>             "clock":1485221821},
>>         {
>>             "host":"precise64",
>>             "key":"lvs_service[inpps,10.211.55.11,7778]",
>>             "value":"0",
>>             "clock":1485221821}],
>>     "clock":1485221821}'
>>  16152:20170124:093701.030 In recv_agenthistory()
>>  16152:20170124:093701.030 In process_hist_data()
>>  16152:20170124:093701.030 In process_mass_data()
>>  16152:20170124:093701.031 End of process_mass_data()
>>  16152:20170124:093701.031 End of process_hist_data():SUCCEED
>>  16152:20170124:093701.031 In zbx_send_response()
>>  16152:20170124:093701.031 zbx_send_response()
>> '{"response":"success","info":"processed: 0; failed: 2; total: 2;
>> seconds spent: 0.000102"}'
>
>
> How to debug LLD issue?
>
> Thanks,
> Linbo
>
>
> ------------------------------------------------------------
> ------------------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, SlashDot.org! http://sdm.link/slashdot
> _______________________________________________
> Zabbix-users mailing list
> Zab...@li...
> https://lists.sourceforge.net/lists/listinfo/zabbix-users
>
>


-- 
С уважением,
Логинов М.А.

mcs...@gm...
+7 (915) 2493122

[Zabbix-users] Debug LLD with trapper item prototype

[linbo l. <ll...@gm...>]

Hi,

In my test environment I try to setup lvs monitor with zabbix LLD.

Zabbix 3.2.0
OS: Ubuntu 14.04.1 LTS

*# Discovery rule*
1. key: lvs.service
2. Type: Zabbix agent
3. Two Filters with empty regular expression
    {#SERVICE_HOST}
    {#SERVICE_PORT}

# cat /etc/zabbix/zabbix_agentd.conf.d/userparameter_lvs.conf
> UserParameter=lvs.service, python
> /etc/zabbix/script/lvs_monitor/dump_service.py
>

# python /etc/zabbix/script/lvs_monitor/dump_service.py
> {"data": [{"{#SERVICE_HOST}": "10.211.55.12", "{#SERVICE_PORT}": "7777"},
> {"{#SERVICE_HOST}": "10.211.55.11", "{#SERVICE_PORT}": "7778"}]}


*# Item prototypes 1 works fine*
  Key: net.tcp.service[tcp,{#SERVICE_HOST},{#SERVICE_PORT}]
  Type: Zabbix agent

# *Item prototypes 2  without zabbix result*
  Key:  lvs_service[inbps,{$SERVICE_HOST},{$SERVICE_PORT}]
  Type: Zabbix trapper

A crontab use zabbix_sender to send data to zabbix_server

# cat /tmp/lvs_service_stat
> precise64 lvs_service[inpps,10.211.55.12,7777] 1485221745 0
> precise64 lvs_service[inpps,10.211.55.11,7778] 1485221745 0
>




*The Zabbix Server log*
>
> trapper got '{
>     "request":"sender data",
>     "data":[
>         {
>             "host":"precise64",
>             "key":"lvs_service[inpps,10.211.55.12,7777]",
>             "value":"0",
>             "clock":1485221821},
>         {
>             "host":"precise64",
>             "key":"lvs_service[inpps,10.211.55.11,7778]",
>             "value":"0",
>             "clock":1485221821}],
>     "clock":1485221821}'
>  16152:20170124:093701.030 In recv_agenthistory()
>  16152:20170124:093701.030 In process_hist_data()
>  16152:20170124:093701.030 In process_mass_data()
>  16152:20170124:093701.031 End of process_mass_data()
>  16152:20170124:093701.031 End of process_hist_data():SUCCEED
>  16152:20170124:093701.031 In zbx_send_response()
>  16152:20170124:093701.031 zbx_send_response()
> '{"response":"success","info":"processed: 0; failed: 2; total: 2; seconds
> spent: 0.000102"}'


How to debug LLD issue?

Thanks,
Linbo

[Zabbix-users] SNMP Triggers

[John W. <joh...@ya...>]

I've got Zabbix setup to monitor an F5 load balancer via SNMP the following parameters are setup on the F5:
Username: adminAuth type: SHAAuth passwd: PasswordXXX123Encryption protocol: AESEncryption passwd: PasswordYYY987
I can do a full SNMPWalk V1, V2, or V3 from the zabbix host to the F5.
I have the following Triggers configured in Zabbix for the F5 Template group:
Severity  Name                   Expression                                                                     Status Disaster sysInterfaceSpeed.1 {Template SNMP Interfaces SNMPv3:int.speed.1.last()}<>1 EnabledDisaster sysInterfaceMTU.3 {Template SNMP Interfaces SNMPv3:int.MTU.3.last()}<1400 EnabledDisaster sysInterfaceMTU.2 {Template SNMP Interfaces SNMPv3:int.MTU.2.last()}<1400 EnabledDisaster sysInterfaceMTU.1 {Template SNMP Interfaces SNMPv3:int.MTU.1.last()}<1400 EnabledDisaster sysInterfaceEnabled.3 {Template SNMP Interfaces SNMPv3:int.enabled.3.last()}<>1 EnabledDisaster sysInterfaceEnabled.2 {Template SNMP Interfaces SNMPv3:int.enabled.2.last()}<>1 EnabledDisaster sysInterfaceEnabled.1 {Template SNMP Interfaces SNMPv3:int.enabled.1.last()}<>1 EnabledAverage Number Connections > 100 {Template SNMP Interfaces SNMPv3:new.conn.last()}>100 EnabledAverage Number-2 connections >100 {Template SNMP Interfaces SNMPv3:new.conn2.last()}>100 EnabledDisaster Interface status:3 {Template SNMP Interfaces SNMPv3:int.stat.3.last()}=1 EnabledDisaster Interface status:2 {Template SNMP Interfaces SNMPv3:int.stat.2.last()}=1 EnabledDisaster Interface status:1 {Template SNMP Interfaces SNMPv3:int.stat.1.last()}=1 EnabledDisaster int.packet.errors {Template SNMP Interfaces SNMPv3:int.packet.errors.1.last()}=1 Enabled

Now I'm trying to pull up some graphs of the SNMP F5 data on Zabbix.  I see none.  So question is, how can I know for sure that Zabbix SNMP monitoring of the F5 applicance is working.  There are no Traps setup anywhere.  Only the Triggers in Zabbix.
Do I need Traps for this to work?
What do I have to see SNMP data in Zabbix?
TIA

Re: [Zabbix-users] Issues with calculated item

[Nicolas C. <za...@nr...>]

Le 03/01/2017 à 20:34, Vinicius Lehmann a écrit :
> Hi,
>
> I want to sum the values ​​of two items with item calculated using the
> last function. In case the values ​​are collected via SNMP and the
> values ​​are ifInOctets. I have implemented a link aggregation switch
> and would like to measure the bandwidth usage of that link aggregation.
> The sum of values does not beat. Help me please.

Hello,

Your calculated item is stored as a "delta", did you tried storing it 
"as it"?

Your ifInOctets items are stored as "delta" because they are counters of 
octets but the final value is bps.

Your calculated item is a simple sum of bps, not the sum of counters.

[Zabbix-users] Issues with calculated item

[Vinicius L. <vin...@gm...>]

Hi,

I want to sum the values ​​of two items with item calculated using the last
function. In case the values ​​are collected via SNMP and the values ​​are
ifInOctets. I have implemented a link aggregation switch and would like to
measure the bandwidth usage of that link aggregation. The sum of values
does not beat. Help me please.

[image: Imagem inline 1]
[image: Imagem inline 2]
[image: Imagem inline 3]
[image: Imagem inline 4]

-- 
Vinicius da Silva Lehmann
*Analista de TI - City10*
*MCP - Microsoft Certified Professional*

Re: [Zabbix-users] v3.2.2 upgrade fails

[Nicolas C. <za...@nr...>]

Le 13/12/2016 à 13:40, Nicolas C. a écrit :
> Hello,
>
> On a Debian Jessie server, after an update from v3.2.1 to 3.2.2,
> "zabbix_server" fails during its startup (log enclosed).
>
> Any idea why ?

It's fixed in v3.2.3 (not tested yet):

http://www.zabbix.com/rn3.2.3

"fixed possible crash when polling vmware.hv.health.status", 
https://support.zabbix.com/browse/ZBX-11573

Re: [Zabbix-users] v3.2.2 upgrade fails

[Volker F. <vol...@gm...>]

This issue has been reported to the bug tracker already. A user on 
#zabbix-de claimed it would start on the second attempt.

Volker

Am 2016-12-13 um 13:40 schrieb Nicolas C.:
> Hello,
>
> On a Debian Jessie server, after an update from v3.2.1 to 3.2.2,
> "zabbix_server" fails during its startup (log enclosed).
>
> Any idea why ?
>
> Thanks for your help,
>
> Nicolas
>
>
> ------------------------------------------------------------------------------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, SlashDot.org! http://sdm.link/slashdot
>
>
>
> _______________________________________________
> Zabbix-users mailing list
> Zab...@li...
> https://lists.sourceforge.net/lists/listinfo/zabbix-users
>

[Zabbix-users] Errors with vmware

[Roy S. K. <ro...@ka...>]

Hi all

I'm trying to figure out how to monitor/discover things with vmware, so I've added some hosts and their macros etc, but I get this in the logs. Checking discovery shows 'not supported' on these hosts, the red info/alert box gives the same as below error message.

I'm currently testing with the root account, but whenever this works, I'd like to use a separate accout without permissions to everything. I can't find any info on that either, so any help is appreciated.

thank

roy

 18793:20161213:113123.357 item "my.servername.tld:vmware.version[{$URL}]" became not supported:
Error returned by expat parser: not well-formed (invalid token)

while parsing serialized value of type string
at line 1, column 409

while parsing call information for method Login
at line 1, column 274

while parsing SOAP body
at line 1, column 264

while parsing SOAP envelope
at line 1, column 38

while parsing HTTP request for method login
on object of type vim.SessionManager
at line 1, column 0

Vennlig hilsen / Best regards

roy
--
Roy Sigurd Karlsbakk
(+47) 98013356
http://blogg.karlsbakk.net/
GPG Public key: http://karlsbakk.net/roysigurdkarlsbakk.pubkey.txt
--
Da mihi sis bubulae frustrum assae, solana tuberosa in modo Gallico fricta, ac quassum lactatum coagulatum crassum. Quod me nutrit me destruit.

[Zabbix-users] v3.2.2 upgrade fails

[Nicolas C. <za...@nr...>]

Hello,

On a Debian Jessie server, after an update from v3.2.1 to 3.2.2, 
"zabbix_server" fails during its startup (log enclosed).

Any idea why ?

Thanks for your help,

Nicolas

Re: [Zabbix-users] "Favourite graphs" widget is missing from the dashboard

[Nicolas C. <za...@nr...>]

Le 20/11/2016 à 11:36, Nicolas C. a écrit :
> Hello,
>
> On Zabbix 3.2, the "Favourite graphs" widget is missing from my
> dashboard. How can I get it back ?
>
> http://i.imgur.com/b6ulMlX.png
>
> I usually use the interface in English but I had to temporary switch to
> French to have the "correct" date format on graphs. Since then the
> widget is gone, even after switching the interface back to English.

Ok, I "solved" it by destroying & re-creating my user.

Re: [Zabbix-users] "Favourite graphs" widget is missing from the dashboard

[Nicolas C. <za...@nr...>]

Le 20/11/2016 à 11:36, Nicolas C. a écrit :
> Hello,
>
> On Zabbix 3.2, the "Favourite graphs" widget is missing from my
> dashboard. How can I get it back ?
>
> http://i.imgur.com/b6ulMlX.png
>
> I usually use the interface in English but I had to temporary switch to
> French to have the "correct" date format on graphs. Since then the
> widget is gone, even after switching the interface back to English.

Note: in "Debug" mode, the widget is there.

******************** Script profiler ********************
Total time: 0.268398
Total SQL time: 0.11479
SQL count: 43 (selects: 29 | executes: 14)
Peak memory usage: 5.75M
Memory limit: 128M

1. drule.get [CControllerDashboardView.php:53]

Parameters:
Array
(
     [output] => Array
         (
         )

     [filter] => Array
         (
             [status] => 0
         )

     [limit] => 1
)
	Result:
Array
(
     [0] => Array
         (
             [druleid] => 5
         )

)

2. graph.get [blocks.inc.php:57]

Parameters:
Array
(
     [output] => Array
         (
             [0] => graphid
             [1] => name
         )

     [selectHosts] => Array
         (
             [0] => hostid
             [1] => name
         )

     [expandName] => 1
     [graphids] => Array
         (
             [28755] => 28755
             [29503] => 29503
             [30169] => 30169
             [57353] => 57353
         )

     [preservekeys] => 1
)
	Result:
Array
(
     [28755] => Array
         (
             [graphid] => 28755
             [name] => HSS : monitor task
             [hosts] => Array
                 (
                     [0] => Array
                         (
                             [hostid] => 10157
                             [name] => tina-hss1
                         )

                 )

         )

     [29503] => Array
         (
             [graphid] => 29503
             [name] => Queue : mx1 + mx2
             [hosts] => Array
                 (
                     [0] => Array
                         (
                             [hostid] => 10329
                             [name] => mx1
                         )

                     [1] => Array
                         (
                             [hostid] => 10488
                             [name] => mx2
                         )

                 )

         )

     [30169] => Array
         (
             [graphid] => 30169
             [name] => Extents pool HYBRID
             [hosts] => Array
                 (
                     [0] => Array
                         (
                             [hostid] => 11851
                             [name] => pmf-nas-mgmt
                         )

                 )

         )

     [57353] => Array
         (
             [graphid] => 57353
             [name] => ops for SAS-RAID10
             [hosts] => Array
                 (
                     [0] => Array
                         (
                             [hostid] => 11851
                             [name] => pmf-nas-mgmt
                         )

                 )

         )

)


SQL (0.000711): SELECT p.value_id,p.source FROM profiles p WHERE 
p.userid=3 AND p.idx='web.favorite.graphids' ORDER BY p.profileid
zabbix.php:21 → require_once() → ZBase->run() → ZBase->processRequest() 
→ CController->run() → CControllerDashboardView->doAction() → 
getFavouriteGraphs() → getFavouriteGraphsData() → CFavorite::get() → 
DBselect() in include/classes/user/CFavorite.php:56

SQL (0.000101): BEGIN
zabbix.php:21 → require_once() → ZBase->run() → ZBase->processRequest() 
→ CController->run() → CControllerDashboardView->doAction() → 
getFavouriteGraphs() → getFavouriteGraphsData() → 
CFrontendApiWrapper->get() → CApiWrapper->__call() → 
CFrontendApiWrapper->callMethod() → CApiWrapper->callMethod() → 
CFrontendApiWrapper->callClientMethod() → CLocalApiClient->callMethod() 
→ DBstart() → DBexecute() in include/db.inc.php:258

SQL (0.000348): SELECT usrgrpid FROM users_groups WHERE userid='3'
zabbix.php:21 → require_once() → ZBase->run() → ZBase->processRequest() 
→ CController->run() → CControllerDashboardView->doAction() → 
getFavouriteGraphs() → getFavouriteGraphsData() → 
CFrontendApiWrapper->get() → CApiWrapper->__call() → 
CFrontendApiWrapper->callMethod() → CApiWrapper->callMethod() → 
CFrontendApiWrapper->callClientMethod() → CLocalApiClient->callMethod() 
→ call_user_func_array() → CGraph->get() → getUserGroupsByUserId() → 
DBselect() in include/perm.inc.php:304

SQL (0.0023): SELECT   g.graphid,g.name FROM graphs g WHERE NOT EXISTS 
(SELECT NULL FROM graphs_items gi,items i,hosts_groups hgg LEFT JOIN 
rights r ON r.id=hgg.groupid AND r.groupid='14' WHERE 
g.graphid=gi.graphid AND gi.itemid=i.itemid AND i.hostid=hgg.hostid 
GROUP BY i.hostid HAVING MAX(permission)<'2' OR MIN(permission) IS NULL 
OR MIN(permission)=0) AND NOT EXISTS (SELECT NULL FROM items 
i,hosts_groups hgg LEFT JOIN rights r ON r.id=hgg.groupid AND 
r.groupid='14' WHERE g.ymin_type=2 AND g.ymin_itemid=i.itemid AND 
i.hostid=hgg.hostid GROUP BY i.hostid HAVING MAX(permission)<'2' OR 
MIN(permission) IS NULL OR MIN(permission)=0) AND NOT EXISTS (SELECT 
NULL FROM items i,hosts_groups hgg LEFT JOIN rights r ON 
r.id=hgg.groupid AND r.groupid='14' WHERE g.ymax_type=2 AND 
g.ymax_itemid=i.itemid AND i.hostid=hgg.hostid GROUP BY i.hostid HAVING 
MAX(permission)<'2' OR MIN(permission) IS NULL OR MIN(permission)=0) AND 
g.graphid IN ('28755','29503','30169','57353') AND g.flags IN ('0','4')
zabbix.php:21 → require_once() → ZBase->run() → ZBase->processRequest() 
→ CController->run() → CControllerDashboardView->doAction() → 
getFavouriteGraphs() → getFavouriteGraphsData() → 
CFrontendApiWrapper->get() → CApiWrapper->__call() → 
CFrontendApiWrapper->callMethod() → CApiWrapper->callMethod() → 
CFrontendApiWrapper->callClientMethod() → CLocalApiClient->callMethod() 
→ call_user_func_array() → CGraph->get() → DBselect() in 
include/classes/api/services/CGraph.php:293

SQL (0.000242): SELECT i.hostid,gi.graphid,h.host FROM graphs_items 
gi,items i,hosts h WHERE gi.itemid=i.itemid AND i.hostid=h.hostid AND 
1=0 ORDER BY gi.sortorder
zabbix.php:21 → require_once() → ZBase->run() → ZBase->processRequest() 
→ CController->run() → CControllerDashboardView->doAction() → 
getFavouriteGraphs() → getFavouriteGraphsData() → 
CFrontendApiWrapper->get() → CApiWrapper->__call() → 
CFrontendApiWrapper->callMethod() → CApiWrapper->callMethod() → 
CFrontendApiWrapper->callClientMethod() → CLocalApiClient->callMethod() 
→ call_user_func_array() → CGraph->get() → 
CMacrosResolverHelper::resolveGraphNameByIds() → DBselect() in 
include/classes/macros/CMacrosResolverHelper.php:448

SQL (0.000517): SELECT gi.graphid,i.hostid FROM graphs_items gi,items i 
WHERE gi.graphid IN ('28755','29503','30169','57353') AND gi.itemid=i.itemid
zabbix.php:21 → require_once() → ZBase->run() → ZBase->processRequest() 
→ CController->run() → CControllerDashboardView->doAction() → 
getFavouriteGraphs() → getFavouriteGraphsData() → 
CFrontendApiWrapper->get() → CApiWrapper->__call() → 
CFrontendApiWrapper->callMethod() → CApiWrapper->callMethod() → 
CFrontendApiWrapper->callClientMethod() → CLocalApiClient->callMethod() 
→ call_user_func_array() → CGraph->get() → CGraph->addRelatedObjects() → 
CGraphGeneral->addRelatedObjects() → DBselect() in 
include/classes/api/services/CGraphGeneral.php:370

SQL (0.000314): SELECT   h.hostid,h.name FROM hosts h WHERE h.flags IN 
(0,4) AND h.hostid IN ('10157','10329','10488','11851') AND h.status IN 
(0,1,3)
zabbix.php:21 → require_once() → ZBase->run() → ZBase->processRequest() 
→ CController->run() → CControllerDashboardView->doAction() → 
getFavouriteGraphs() → getFavouriteGraphsData() → 
CFrontendApiWrapper->get() → CApiWrapper->__call() → 
CFrontendApiWrapper->callMethod() → CApiWrapper->callMethod() → 
CFrontendApiWrapper->callClientMethod() → CLocalApiClient->callMethod() 
→ call_user_func_array() → CGraph->get() → CGraph->addRelatedObjects() → 
CGraphGeneral->addRelatedObjects() → CHost->get() → DBselect() in 
include/classes/api/services/CHost.php:432

SQL (9.6E-5): COMMIT
zabbix.php:21 → require_once() → ZBase->run() → ZBase->processRequest() 
→ CController->run() → CControllerDashboardView->doAction() → 
getFavouriteGraphs() → getFavouriteGraphsData() → 
CFrontendApiWrapper->get() → CApiWrapper->__call() → 
CFrontendApiWrapper->callMethod() → CApiWrapper->callMethod() → 
CFrontendApiWrapper->callClientMethod() → CLocalApiClient->callMethod() 
→ DBend() → DBcommit() → DBexecute() in include/db.inc.php:319