Zabbix queue too high?

[Deeprain]

Hello together,

I'm currently setting up a zabbix setting with one master and two active proxies.
At the proxies, I'm monitoring 200 hosts with in sum approximately 10000 items.

All server, proxies and agents have the recent version 3.0.11. But this problem was also in version 3.0.10.

So far to the environment. Now I stumbled across the zabbix internal performance measurements and I got the following results for the zabbix queue (screenshot2).

As you can see, there are a lot of items currently in the zabbix queue at the master.
Of course, the number as well as the concrete items in the queue change over time
(usually items are for 10s to 1min in the queue), shown in screenshot1.

But seen overall, there is always quite a high number of items in the queue.

So I'm wondering, whether this number is common for this monitoring environment?? In my gut feeling, it seems too high for me, but I'm not really sure.
If it's too high: where are actually the values or things I may have a look at or I have to tune??
Is increasing the pollers an option, or better: will this help to lower the queue items?
Or is there any other option, as I think, the queue should be lower. In some forums or
just googling around, I saw sometimes the statement "the queue should*trend towards zero", which is obviously not the case in my setting.

What do you think: do I have a problem or not?

Thanks and kind regards,
Deep

[Clontarf[X]]

For active checks, we have found the queue seems to be consisting of items that are waiting for a sample, not neccessarily that there is any process "waiting" or "hanging".

What would be extremely handy to have is the ability to see the *entire* queue, as opposed to only 500 items.

For example, our Active queue is anywhere from 40000 to 70000 items, due to the nature of our infrastucture.