Gaps on all my Graphs - Help

[chris_dinis]

Gaps on all my Graphs - Help

I’m hoping someone can advise me; yesterday I upgraded zabbix from 1.0 to 1.1. and now I have gaps on all my graphs as per attached pic.

what could be causing this.

I have looked at http://www.zabbix.com/forum/showthread.php?t=1868
and
http://www.zabbix.com/forum/showthread.php?t=1364. (no good because 1.1 does not include include/classes.inc.php or should it?)

Any ideas welcome

Regards

chris

[just2blue4u]

try increasing the amount of memory, php may use (in php.ini)...

[Alexei]

The problem does not have anything to do with memory settings in php.ini.

How many unreachable hosts you have?

[nick5000]

What is your CPU load?

[SLMitch]

I have also enormous gap when I use over a one maybe two hours period.

There are none UNKNOW value.

The solution i found is ... not use short period.

[chris_dinis]

The problem does not have anything to do with memory settings in php.ini.

How many unreachable hosts you have?

Alexei, if i look on the zabbix_server.log file it shows me no unreachable hosts.

I know with 1.0 if I stopped zabbix for 10mins and then restarted, it would fill the gaps on the graph, could it be a configuration setting that will allow the gaps to be filled.


And my CPU is running very low (between 0.6-25% at it worst day, avg 1.81%)


SLMitch, your solution is not an acceptable solution. Zabbix is better than that!

Thanks

[scott]

Another variation

I experience the same problem whenever I modify a Trigger for a specific item. The graph for the item will have a gap of a few minutes.

[pdwalker]

I have also enormous gap when I use over a one maybe two hours period.

There are none UNKNOW value.

The solution i found is ... not use short period.

I've seen this before, I'd be curious to see what your values were

How often were you polling the data? What did you change the value to?

Thanks

- Paul

[Nate Bell]

There are none UNKNOW value.

Are there gaps in the data itself? I've seen these gaps in graphs happen when the Zabbix server or agent gets bogged down and the data doesn't get collected. Look at your data as a list of text values with timestamps and see if there are gaps of time where data was simply not collected. I've had this happen a few times before, but usually while tweaking Zabbix settings, or when a CPU intensive program on a monitored server screws up the queue.

Nate

[pdwalker]

Are there gaps in the data itself? I've seen these gaps in graphs happen when the Zabbix server or agent gets bogged down and the data doesn't get collected. Look at your data as a list of text values with timestamps and see if there are gaps of time where data was simply not collected. I've had this happen a few times before, but usually while tweaking Zabbix settings, or when a CPU intensive program on a monitored server screws up the queue.

Nate

I think you will find there are no gaps in the data. There wasn't when my graphs had the same problem.

- Paul

[chris_dinis]

I found that if I increase the Update interval from 30 to 40 secs, it fixes most of the problems! Though I still get an occasional gap (once an hour or so)
My question is what has changed from zabbix 1.0 and zabbix 1.1, why did it work on version 1.0 and not on 1.1.

I've attached a new graph of what it looks now; and you can compare the difference of what it looked like at first after the upgrade.

Pdwalker – you are right their was no gaps in the data.

Any suggestions on what should be the lowest Update interval time on a normal query (like bandwidth)?

[pdwalker]

Try 60 seconds.

For my own servers, I monitor the ethernet traffic value every 30 seconds without gaps in the graph. Anything below this causes the "gaps in the graph" problem for me.

At a guess (I don't know because I have not tried to track it down) I guess it has something to do with some minor bug in the library used to create the graphs.

- Paul

[drathi81]

Hi,

I'd like to bring back this thread on topic.

I've installed Version 1.1.1 on the following machine:
Dual Intel(R) Xeon(TM) CPU 3.20GHz
2GByte RAM
2x70GByte HDD (RAID-1)

At the moment we're monitoring about 8576 itmes on 16 hosts (linked via template). The moment I put in more hosts, we start getting gaps in our graphs AND in our data-collection. About 40% of the items are updated every 90sec and the others every 60sec.
When I look with top on my system usage I get only about 20%CPU-usage and about 98% memeroy-usage (swap is ununsed). I've also monitored the port the zabbix-machine is connected to and I've seen traffic not rising over 1Mbit on a GBit-Connection.

Could anybody explain me, why I get these gaps in my data collection (about 5 minutes of no data capturing) even if the machine and the network-port seem not to be the limiting factor?


Regards,
drathi81

[sbadger]

How long does it take to proccess your queue? Are items sitting out there in the 5 minute or > 5 minutes queues?

[stever]

I had a similar problem. Try tweaking the StartPollers and StartTrappers parameters in zabbix_server.conf. I have a very fast machine (dual opteron server), and saw no load on the server, but my queue had a ton of 5 minute + entries. Now with 30 pollers (probably overkill), everything happens instantly. Again, the Monitoring -> Queue tab should tell you what your problem is.