The eternal IT-services

[Alexei]

I am sorry for the lack of useful information from our side. I expect the issue will be verified and fixed (if it needs to) prior to release of 1.4.5. We are really busy these days. You will see result of our work in 1.4.5 and 1.5 very soon.

[noxis]

I am sorry for the lack of useful information from our side. I expect the issue will be verified and fixed (if it needs to) prior to release of 1.4.5. We are really busy these days. You will see result of our work in 1.4.5 and 1.5 very soon.

Thanks for the update Alexei, is it viewable in the 1.4.x trunk?

[qix]

Thanks for the feedback Alexei.
I await the next release with great anticipation

[skogan]

Thanks, Alexei, finally some good news. While you're at it, can you also make the IT services calculation monthly instead of weekly? Or, better yet, selectable

[Niels]

So, what's the status of this after two months?

[Alexei]

Well, I did number of tests and everything seemed to be working. I think that main issue is the documentation, currently there is almost no detailed information about IT Service in the manual.

[hardtofi]

Well, I did number of tests and everything seemed to be working. I think that main issue is the documentation, currently there is almost no detailed information about IT Service in the manual.

You mean if you do like this:

1. Set up two triggers
3. Set up an it service with two children showing "min children", one depending on each of the two triggers
4. wait

Then, is the parent showing the min of the children and are the children showing the correct values for you?

[bee]

I am sorry for the lack of useful information from our side. I expect the issue will be verified and fixed (if it needs to) prior to release of 1.4.5. We are really busy these days. You will see result of our work in 1.4.5 and 1.5 very soon.

Hi @,
how does IT services for you after ZABBIX 1.4.5 release?

For me, it worked!

Thanks,
BEE

[exkg]

Hi,

We cant use IT Services yet ... but I have a question:

- possible is use 'Applications' in IT Services ?



[]s,
Luciano

[richlv]

Well, I did number of tests and everything seemed to be working. I think that main issue is the documentation, currently there is almost no detailed information about IT Service in the manual.

while that might be true, i have a hard time understanding why a trigger that is flipping almost daily gets 100% sla graph (both last seven days & yearly graph when clicking on a graph for that particular trigger) and i can see it flapping in the events page.

the same also happens if sla is measured for an item that is disabled - instead of showing unknown, it's all green...

there also are several reports :

https://support.zabbix.com/browse/ZBX-366

https://support.zabbix.com/browse/ZBX-384

https://support.zabbix.com/browse/ZBX-141

[marcis]

Just to share my experience.

Don't know the cause, but I needed to find a solution for this "100% OK" thing while actual trigger history shows a lot of downtime.

This is connected with my other post http://www.zabbix.com/forum/showpost...15&postcount=2

I had to provide a monthly sla report for about 800 hosts when it turned out that something was wrong with SLA calculation. For some months SLA was not calculated and showed "100% OK" and it was a lie. When looking at database I found that table "service_alarms" was definitely missing a lot of entries. Thankfully the table "events" holds approximately the same information (if I get it right), so I just did some bash scripting and sql querying.

What I did was basically copy entries from "events" to "service_alarms" translating from "triggerid" (or "objectid", but from php code I've understood that they're basically the same) to serviceid and later replacing all values "1" (true in "events") to "4" ("problem" in "IT services") and left "0" as they were. The entries from "events" whose values was "2" I simply ignored because they mean status "unknown".

During this process I couldn't figure out why xxxxxxid's of tables are not set to "auto_increment", but instead are managed with help of table "ids". Of course four my affair I temporarily set the "auto_increment" flag and later updated table "ids".

Hope i'll give some Ideas for others. I would be glad to share more detailed information if there's interest in it.

p.s.

the reason why SLA's was showing 100% remains unclear. Maybe I was the one who damaged them, but for some time they were working fine. I can't be sure that I'm not the one to blame because all SLA table was created automatically from hosts already added to system .... and the hosts we're also imported by some magic from "Cacti".

I wouldn't say that SLA's are broken ... maybe more documentation and some troubleshooting guide would be needed

[richlv]

it looks like revision 5952 on trunk has finally fixed availability report graphs (reports -> availability report -> 'show').

opening similar graph from it services also finally seems to actually draw red areas for failed checks, though it seems that historical data is ignored for newly created services and is calculated only for the preiod while service has been active (and all period in the past is reported as ok, even though there is no data at all).

so availability report graphs finally show more or less correct data (and does not draw green lines for periods where there is no data at all);
sla/services calculation draws green/ok data for periods without data and is not able to re-use historical data for sla calculation.

in my trunk test installation i have a (seemingly) normal trigger availability graph, and a large green sla graph with a tiiiiny red line

[banderas20]

I have the release 1.6.5 and I still got the same error.

¿any new issues?

[richlv]

I have the release 1.6.5 and I still got the same error.

same error being ? this thread mentioned several issues

[banderas20]

same error being ? this thread mentioned several issues

"IT Services always at 100%"

As stated here:

http://www.zabbix.com/forum/showthread.php?t=8129

http://www.zabbix.com/forum/showthread.php?t=13054