I get the zabbix_agentd.conf from Zabbix 1.4.7 sources (the compiled agent doesn't have the .conf). In the zabbix_agentd.conf I don't see any particular configuration about user and user group
Ad Widget
Collapse
AS/400 Monitoring solutions
Collapse
X
-
-
Hi phdeliege,
my Zabbix Agent emulator uses the QYASPOL system API call (link) for both disks (using YASP0300 format) and ASP's (using YASP0200 and YASP0100 format) statistics.
According to documentation, this API returns the capacities in megabytes, where:
So, this Agent multiplies all these values onto one million (1000000) to return a volumes in a bytes.Megabytes, as used in this topic, means millions of bytes.
Units for Zabbix items for all these values is set as "B" (bytes), so Zabbix uses traditional for IT 1024-based notation (link); more exactly, MiB - mebibytes and GiB - gibibytes (link).
Probably, original IBM utilities uses 1000-based transformation (but 1024-based).
So, really (142.22 GiB) = (142.22 * 1024 * 1024 * 1024 bytes) = (152707562209.28 bytes) ≈ (152.7 GB)
And: (71.11 GiB) = (71.11 * 1024 * 1024 * 1024 bytes) = (76353781104.64 bytes) ≈ (76,354 MB)
Comment
-
ikarisroy, all versions of my Zabbix Agent emulator for AS/400 were published with a source code, example of config-file, documentation and a compiled jar-file.
The example of config-file really does not include the "User=" parameter (as usually it is not required), but the documentation has note about it in the section "Start of program".
Unfortunately, the last year the "Zabbix Share" site became unavailable, so the last version of my project (0.7.8 at the moment) is still temporary available at my Google Drive (see for a correct link to this message).
Probably, it will be migrated onto GitHub; but I can't promise exact terms.
Last edited by Kos; 01-03-2022, 17:34.Comment
-
Hello everyone, hope you doing good,
I'm new in monitoring and i start using zabbix, i found it usefull and i want to monitor microsoft sql server using zabbix but whene i add the MSSQL BY ODBC themplate i didn't get any information my server, and everything is configured right i'v trying a week now with no solution if you can help,,,,, i'm using zabbix server on debian, and ms sql server is installed on windows server 2012 r2, the connexion between zabbix server and windows server work successfully using zabbix agent or snmp.
Help pls asapComment
-
thiomed, probably, you need to open a new thread for your specific topic (about connectivity with MS SQL server and the appropriate template usage).Hello everyone, hope you doing good,
I'm new in monitoring and i start using zabbix, i found it usefull and i want to monitor microsoft sql server using zabbix but whene i add the MSSQL BY ODBC themplate i didn't get any information my server, and everything is configured right i'v trying a week now with no solution if you can help,,,,, i'm using zabbix server on debian, and ms sql server is installed on windows server 2012 r2, the connexion between zabbix server and windows server work successfully using zabbix agent or snmp.
Help pls asap
This thread is about different topic.Comment
-
-
Hi, I'm looking for some help with a new installation, please.
In this configuration
### Option: ServerActive
# List of comma delimited IP port (or hostname port) pairs of Zabbix servers for active checks.
# If port is not specified, default port is used.
# If port is not specified, square brackets for IPv6 addresses are optional.
# If this parameter is not specified, active checks are disabled.
# Example: ServerActive=127.0.0.1:20051,zabbix.domain
#
# Mandatory: no
# Default:
ServerActive=172.18.221.24:10050
And Zabbix items are set: Zabbix agent (active)
Then the Agent stops after some minutes, and no Zabbix agent (active) results are displayed on the Zabbix server
However, in this configuration
### Option: ServerActive
# List of comma delimited IP port (or hostname port) pairs of Zabbix servers for active checks.
# If port is not specified, default port is used.
# If port is not specified, square brackets for IPv6 addresses are optional.
# If this parameter is not specified, active checks are disabled.
# Example: ServerActive=127.0.0.1:20051,zabbix.domain
#
# Mandatory: no
# Default:
# ServerActive=172.18.221.24:10050
With Zabbix items set: Zabbix agent
Then the Agent remains active, and most Zabbix agent results are displayed on the Zabbix server.
(but not all as it appears some (MSGQ) must be in Zabbix agent (active) mode)
Extract of Log file when ActiveChecks allowed:
30:20220609:071534.918 End of send()
30:20220609:071534.919 active check configuration update from [172.18.221.24:10050] started to fail (java.net.SocketException:
30:20220609:071534.920 end of refreshActiveChecks(): false
30:20220609:071534.920 in processActiveChecks() server:'172.18.221.24' port:10050
30:20220609:071534.921 End of processActiveChecks()
The defined port 10050 has been allowed both directions (IBMi / Zabbix)
Basically, the Agent stops and no results are passed to Zabbix server in ActiveChecks mode.
Has anyone had the same?
Last edited by dm_bry; 09-06-2022, 07:34.Comment
-
Kos, since years I´m monitoring the AS400 , I think you made the Java Version that is running on our machine. Really stable over years now.
I´m running the Java in an own subsystem.
We have upgraded to zabbix 6 version month ago, in general no problem, still using the same java on AS400.
I remember, you helped with some hints years ago.
Since some days I`m trying to use one of the newer features, Custom Intervals on items.
I try to get rid of some triggers coming in the night, where the machine is busy, and e.g. a MSGW is there for some minutes.
Also some monitored process are not running on Sundays.
As soon as I apply this custom intervals the AS400 gets really busy, so instead of 10% CPU all over all processes the zabbix process eats 60 %.
Custom interval: flexible : Interval 0 1-7 22:00-24:00
Well, I`m not that I5 expert, but in the JOB log I can see that the agent trys now every second to start a new process.
Off course this makes the machine very busy... I do not get friends with that in the company...
Did you ever had similar issues?
I´m using zabbix agent v.0.7.7,
Probably I can solve this issue in another way, but maybe a solution is already known.
I have the feeling, that the JAVA program might pickup the flexible interval 0 wrong and is trying to do the job in a very short interval.
Gerald
Comment
-
Hi RohrbaGe,
What exactly metrics do you use with a custom intervals? What mode of checks are used (active or passive)?
Custom intervals are not something specific for AS400.
If you use them for passive checks, then all pollings are performed by Zabbix server (or Zabbix proxy); Zabbix agent just respond onto these polling requests - it should work.
However, if you like to use an active checks (for example, for monitoring of logs that possible in active mode only), then a custom intervals must be supported by Zabbix agent.
Unfortunately, our Zabbix agent emulator does not support custom intervals for an active checks.
I'm not sure how our Agent will process a zero interval value ("0") for an active checks - it's interesting question, I didn't test this. Probably, it could cause to high utilization.
The last version of Agent at the moment is 0.7.8 (see here), but it has no changes in this area.
Just for information: the installation of current PTF on the system i v7.4 causes to this Agent stopped to monitor all ASP's besides the very first one. The system ASP (ASP1) is OK, but all further - not.
I'm aware of this problem and investigating this case.
--
ConstantinComment
-
Constantin,
yes active checks, and yes if the agent does not support this it will end in very high cpu usage.
That confirms my experience.
Well, it's good to know that this won't work, but I think I can change the trigger for date and time ranges I do not want an alarm.
Good hint with he PTF, we still have 7.3 . As we plan to change to a new ERP System not based on I5, our machine will not be upgraded anymore.
But still needs to run 2-3 years somehow. So the dinosaur is going to die..
Regards
Gerald
Comment
-
Bad news: the same effect is present for disks also (not ASP's only). I.e. the very first disk is parsed correctly, all others - wrong. Affected all metrics vfs.fs.* and as400.disk.*.
Good news: I found how to make a workaround.
Bad news: I'll be out of office nearest 2 weeks, so I'll be able to build a new release a bit later.
Good news: a bit later new release will be
Comment
-
Comment