Cant edit down device.

[MrRat]

When a cacti device is down. (ping and/or snmp down). I cannot edit the device in cacti. When clicking the device it just twirls and browser never responds. I have to close browser and re-open to get back to cacti. This has been an issue for me since day 1 I think and have always just dealt with it when a device IP changes I just delete or disable the device and re-add it with new IP. Is there something I can do to allow me to edit down devices in cacti?

[victorantunes]

How long do you usually wait before closing the browser window?

The time it takes me to access downed devices may very between 3 to sometimes around 30 seconds or even more. I've gotten used to it.

The first I'd look at are your server's specs and make sure it's not too busy handling devices to deal with the additional checks so that it ends up taking too long to respond.

[MrRat]

yeah, I've waited up to multiple minutes...

Code: Select all

05/24/2013 04:09:17 PM - SYSTEM STATS: Time:14.9840 Method:spine Processes:1 Threads:50 Hosts:1328 HostsPerProcess:1328 DataSources:44998 RRDsProcessed:0  
05/24/2013 04:08:28 PM - SYSTEM SYSLOG STATS:Time:2.85 Deletes:0 Incoming:2782 Removes:8533 XFers:0 Alerts:18 Alarms:0 Reports:0  
05/24/2013 04:08:25 PM - SYSTEM THOLD STATS: Time:7.8857 Tholds:1846 TotalHosts:1327 DownHosts:12 NewDownHosts:0  
05/24/2013 04:08:17 PM - SYSTEM STATS: Time:15.8939 Method:spine Processes:1 Threads:50 Hosts:1328 HostsPerProcess:1328 DataSources:44997 RRDsProcessed:0  
05/24/2013 04:07:30 PM - SYSTEM SYSLOG STATS:Time:2.8 Deletes:0 Incoming:2553 Removes:8492 XFers:0 Alerts:18 Alarms:0 Reports:0  
05/24/2013 04:07:27 PM - SYSTEM THOLD STATS: Time:10.9052 Tholds:1847 TotalHosts:1327 DownHosts:12 NewDownHosts:0  
05/24/2013 04:07:16 PM - SYSTEM STATS: Time:14.9801 Method:spine Processes:1 Threads:50 Hosts:1328 HostsPerProcess:1328 DataSources:45001 RRDsProcessed:0  
05/24/2013 04:06:30 PM - SYSTEM SYSLOG STATS:Time:3.04 Deletes:0 Incoming:2853 Removes:8872 XFers:0 Alerts:18 Alarms:0 Reports:0  

[nditguy]

I have the same issue. I am only monitoring about about 15 devices. I have waited up to 30 minutes. Any resolutions, ideas?

[gandalf]

Which timeout settings are you using for those hosts?
R.

[MrRat]

even devices with ping timeout of 500 and snmp timeout of 600 this is the case. If the device goes down I cannot edit it without working around this problem. I have figured out that I can check the box for the device and select update availability options, then set it to "none". after that I can edit the device.

[pnyet]

Hi,

I have same problem, when I use nginx, php-fpm and perconadb (mysql), please see attached file for my cacti setup and see below for log from nginx:

2013/08/15 07:08:34 [error] 26090#0: *593272 upstream timed out (110: Connection timed out) while reading response header from upstream, client: 10.0.0.10, server: localhost, request: "GET /host.php?action=edit&id=227 HTTP/1.0", upstream: "fastcgi://127.0.0.1:9000", host: "mrtg.mytrafficinfo.local", referrer: "https://mrtg.mytrafficinfo.local/host.php"

I found that there was timeout happened but when I'm increase the timeout value on php.ini and cgi timeout to 120s or 300s but doesn't solve this problem.

Maybe this problem related with mysql performance?

Thank you

[gandalf]

even devices with ping timeout of 500 and snmp timeout of 600 this is the case. If the device goes down I cannot edit it without working around this problem. I have figured out that I can check the box for the device and select update availability options, then set it to "none". after that I can edit the device.

Are you able to run a tcpdump when trying to edit such a device?
I expect to find a lot of snmpget's but all of them equipped with the timeout defined for your host.
In case I'm wrong, this would explain your findings.
R.

[pnyet]

even devices with ping timeout of 500 and snmp timeout of 600 this is the case. If the device goes down I cannot edit it without working around this problem. I have figured out that I can check the box for the device and select update availability options, then set it to "none". after that I can edit the device.

Are you able to run a tcpdump when trying to edit such a device?
I expect to find a lot of snmpget's but all of them equipped with the timeout defined for your host.
In case I'm wrong, this would explain your findings.
R.

hi Gandalf,

Thanks for your reply, I would like to inform you that the problem only happened on device (down) that have TCP ping detection.
If you're need more information please let me know.

Thank you

[gandalf]

Thx, but this does not help me nailing this down. Still need a tcpdump and perhaps even more
R.