A few questions

[Cray]

Hi all,

Since I'm actually actively working on Zabbix deployment, I have a few questions regarding some Zabbix behavior :

1. I have multiple windows workstations monitored (the data comes trough by proxy).
I have a strange bug (if it is one) regarding one workstation : the data(s) collected is received by the master server (graphs works well), but in the host configuration panel, I see

Available : Not available
Error : Get value from agent failed: ZBX_TCP_READ() failed [Connection reset by peer]

Despite this error message, the workstation is being monitored correctly.


2. Is there any possible way of clearing the discovery history ?
The fact is, I made some upgrades to my discovery rules (after adding proxies), and the old discovery results are still present in the monitoring / discovery panel.

I tried to manually delete the records by excuting the following SQL statements :

DELETE FROM history
DELETE FROM history_str
DELETE FROM history_uint

but it didn't work

Any ideas ?


3. I made some screens, with all the graphs set to dynamic (so that I can see the screen's result for all the similar hosts).
The screen refreshes according to the selected host, but the problem is that I can only click on a particular graph (and see it's details) when the host is set to "default".

Is that a normal behavior



Thanks in advance

[xs-]

1. Try upping either pollers (for passive agent/server based checks) or trappers (for active agents) in the zabbix proxy config.

2. It can be done, you are probably deleting the wrong records. Im not using discovery atm, so i cant tell you which one precisely. Take a closer look at the database layout, column names and content.

3. sounds like a bug, submit it

[Aly]

3. it's not a bug. Already discussed in some thread..

[MrKen]

3. it's not a bug. Already discussed in some thread..

Right here: http://www.zabbix.com/forum/showthread.php?t=11944

[xs-]

3. it's not a bug. Already discussed in some thread..

I stand corrected, its not a bug, dont submit it