AS/400 Monitoring solutions

[sancho]

Hi bigredau15

That value is configured in a macro inherited from the template.
If you change the value of the macro in the template, that change will affect all the hosts that have that template applied.

To modify it only on a host, you must go to the "Macros" section of the host and there you will see the inherited macros where you can modify the threshold of that macro.

I hope that helps.

[baburao]

Appears we can do most all that we need with the exception of channel-dependent on the client name. Our Developers do testing on PROD machines and we have to channel messages produced by the designers. Appears there are no different choices for them to demonstrate this is a test message. Would it be conceivable to affix to the start or end of the string and if clear, either don't tack it on or put something like "username_blank". I will be OOO one week from now, I will check when I return, possibly you will have another form by at that point.

[bigredau15]

Thanks sancho, much appreciated.

Our only remaining issue that we have is subsystem *ACTIVE / *INACTIVE triggers:

Trigger Expression:
{LPAR1:as400.subsystem[QBATCH,QSYS].last()}="*INACTIVE"

Recovery Expression:
{LPAR1:as400.subsystem[QBATCH,QSYS].last()}="*ACTIVE"

The trigger expression creates the problem perfectly fine when the subsystem is *INACTIVE, but when the subsystem becomes *ACTIVE, the problem doesn't disappear.

Even without the Recovery Expression in place, if we close the existing problem for *INACTIVE, the problem comes back even though the subsystem is *ACTIVE on the iSeries LPAR.

Any ideas?

Cheers

[sancho]

Hi,

Try these triggers:

Trigger Expression:
{LPAR1: as400.subsystem [QBATCH, QSYS] .str ("* INACTIVE", # 1)} = 1

Recovery Expression:
{LPAR1: as400.subsystem [QBATCH, QSYS] .str ("* ACTIVE", # 1)} = 1

This monitoring must be text type.

I monitor the subsystems with the key proc.num [name, user, state, subsystem]
With which I can see if the subsystem is booted and the amount of jobs it has.

For example, in your case the key would be something like this:
proc.num [,,, QBATCH]

It will give you a numerical value, if it is 0 it means that it is not started, so the trigger would be something like this:

Trigger Expression:
{LPAR1: proc.num [,,, QBATCH] .last (# 1)} = 0

Recovery Expression:
{LPAR1: proc.num [,,, QBATCH] .last (# 1)} = 1


I hope it helps you.

[bigredau15]

Thanks for the assistance, unfortunately both methods failed to work. I must be doing something wrong.

Is there a possibility for you to take some screenshots of the Item and Trigger page that you've got configured?

Thanks again

[sancho]

Hi bigredau15


Item and trigger key subsystem:




Item and trigger key proc.num:





Tell me if this helps you.

[bigredau15]

Thanks Sancho!

The proc.num method worked - but I have no idea what I did differently today that I did yesterday (when it didn't work)! Either way, thank you for all of your help.

Cheers,

[sancho]

There is nothing to be thankful for, the forum is there to help us.

Another greeting for you.

[kamal123]

It appears to be a bug (or, in any event, some undocumented "include") in API utilized. We get the absolute CPU utilized (in milliseconds) for each activity and contrast it and the past incentive for a similar activity to acquire the CPU use for the last "tick". In principle, the absolute CPU utilization for the activity ought to be expanding as it were. Actually, in some cases, we abruptly get 0 (zero) for a few occupations that before as of now had non-zero all out CPU use. Subsequently, negative qualities are showing up. amit bhar art

[Maran Ganapathy]

Dear All,

When I import the template for AS400 which has been downloaded from the url https://share.zabbix.com/operating-s...agent-emulator

I get the error "Import failed: Cannot import template "Template AS400", linked template "Template App Zabbix Agent" does not exist.

Please help me to resolve this issue.

Warm Regards
Maran

[joshi123]

truly, you can utilize the other machine with Java 7 or Java 8 to begin this Zabbix specialist emulator and associate with the i5 framework by means of an organization. It doesn't utilize a realistic, so X11 isn't important; support access is sufficient. You need, be that as it may, some extra JAR-records (see the documentation for subtleties). amtemu

[Teamster]

I am using the {Host.name} in the trigger description, however I want to display the first 15 characters ONLY. How can I do this?

[camiespico]

Hi Kos! I am using your agent for almost 2 years and its been a great asset for me. Thanks for that! Anyways, I am having trouble with active check message queue items with the eventlog . They were working fine until a couple of weeks ago when suddenly items started to get to a not supported status reporting an MCH2804 not supervised for QGYOLMSG. I read a couple of notes saying that the app is not closing the lists after reading the message queue. I restarted the agent, even upgraded the ZabbixAgent.jar to the latest version but nothing seems to work. I get the messages but keep getting them 5 to 10 minutes after i should.

I saw on your readme that you already fixed this bug in an earlier version of the agent, Could you help me? Thanks!!

[Kos]

Hi Kos! I am using your agent for almost 2 years and its been a great asset for me. Thanks for that! Anyways, I am having trouble with active check message queue items with the eventlog . They were working fine until a couple of weeks ago when suddenly items started to get to a not supported status reporting an MCH2804 not supervised for QGYOLMSG. I read a couple of notes saying that the app is not closing the lists after reading the message queue. I restarted the agent, even upgraded the ZabbixAgent.jar to the latest version but nothing seems to work. I get the messages but keep getting them 5 to 10 minutes after i should.

I saw on your readme that you already fixed this bug in an earlier version of the agent, Could you help me? Thanks!!

Hi camiespico,
according to IBM documentation, this error message really does mean that some list of objects (most probably, a list of messages in the message queue) is regularly not closed. However, it seems that this list is closed correctly by our Agent every time.
Could you like, please, send me a log file of the Agent emulator onto my e-mail mentioned in the README?
This log file should include messages, at least, from the last agent start, and up to the moment when the problem is occurred.

[bobf]

Hi Kos. Don't do much in these sort of forums (old dog new tricks). Anyway we've started to use the Java Agent internally and at some of our clients. I must say I'm quite impressed. Looked at the code and closed it quickly (haven't done anything in java for probably 20 years). Anyway I think I found a couple of "bugs". I've put in some workarounds so nothing I couldn't deal with.

1. If I put something in operational data in a trigger it will output that data and it shows in the problem. All good but if I later take that value out it seems to keep the original value. Not tried it again for a bit but that seems to be the case. Maybe I need to restart the agent.
2. I've tried using flexible custom intervals but found that if I have an update interval on the item with a shorter interval to the one in the schedule it doesn't work but if the interval on the item is greater than the one in the schedule then it does work.
3. If the interval on the item is set to zero and we have a an interval in the schedule - and it happens to generate a trigger (in my case the qinter subsystem was inactive) I got something like 11k records in the history before I could stop it.

Lastly are you planning on looking at interactive response times as one of the metrics later ? Not just the cpu percentage. Are there any plans for another release soon and what may we look forward to ?

Thanks again. Love your work
Bob

[hab]

Hi camiespico,
according to IBM documentation, this error message really does mean that some list of objects (most probably, a list of messages in the message queue) is regularly not closed. However, it seems that this list is closed correctly by our Agent every time.
Could you like, please, send me a log file of the Agent emulator onto my e-mail mentioned in the README?
This log file should include messages, at least, from the last agent start, and up to the moment when the problem is occurred.

Hi, I also see this error on a single partition. Any idea how I can get it to become supported again? I have gone through agent log, joblogs and similar, but I cannot find any errors relating to this error.