Ad blocker detected: Our website is made possible by displaying online advertisements to our visitors. Please consider supporting us by disabling your ad blocker on our website.
Can help on CMDPHP ERROR: Invalid Return Value in ss_host_cpu.php for get 4000 ? i see on https://github.com/Cacti/cacti/issues/1143 was solved but i used 1.2.14 still apear this error suddenly, on first setup no error on this part
When did you create the Graph? It may take some time to appear. Did rrd File for this Graph already created? And what did your log File show you? check #php poller.php under cacti folder and try to rise your log level from low to Medium and watch what will happen in your log file.
When you have problems like this, it's helpful to run spine in debug mode to ensure that the script/snmp is returning the correct data. Like this:
./spine -R -V 3 -f device_id -l device_id -S
Replace "device_id" with your Device ID from the Device Management page. If you are getting good data, then you should look secondarily to maximum values in the RRDfile, and if that fails, look at your Step and Heartbeats and make sure that they align with the Cacti Poller interval.
In other words, if you have your Cron Interval set to 5 minutes, and your poller interval set to 5 minutes, but you are trying to graph at a 1 minute granularity, then Cacti will never work. Much of this is laid out in the documentation.
True understanding begins only when we realize how little we truly understand...
to be frank, I also expirience a behaviour like this in my new installed 1.2.14 cacti too. I had this Problem with 1.2.13 cacti and I upgraded thinking that I may be have to do that. After upgrade i could see things got more worser, that I could not see many devices up they were seen all down in cacti although they were up. I got fed up and I removed the whole MariaDB and started installing new with 1.2.14 and installation went smooth, showing that I am experienced in cacti installing now.
Now, for the first let us say 100 Devices I never had seen an issue the more my Devices increase in Number, the more devices got into a status unknown. And never get up from unknown status since no rrd files were created for them. This time I am smarter to use only cmd.php only. I have spine installed and well configured, but I had those problems in spine that I suspect spine and leave it aside for a while till every thing runs first. Just like my friend laruku54 I only use cmd.php as a poller and not spine.
Some of my devices shows up as having the status UP, which you think is nice, however the Graphs are empy no data is seen. Meaning that either the devices are unknown and have the cactus pic as a graph or they seem to be up but they don't do what we in cacti expect - empy Graphs are kidding us.
What is going on? crontab is well and at least for the first 100 Devices it was perfect. It can also not be that it it is a permission of Folders issue, as we have seen Graphs for devices already. After a while even the Graphs which were created properly, we see them gradually dispaear.
@theWitness: what do you mean by
## In other words, if you have your Cron Interval set to 5 minutes, and your poller interval set to 5 minutes, but you are trying to graph at a 1 minute granularity##
I have read much about cacti and did my google searchin to solve this issue, but unfortunately I can't deliver a satisfyable result to my BOSS, that they are thinking to see around to find a better monitoring software. End I remain nothing but a looser, because I convinced them to stay with cacti we had for many years.
Does cacti have a problem in SUSE SLES 15.1?
Is cacti good only for 100 Devices and no more?
Is there any support I can get from you guys?
2020-10-24 18:20:38 - POLLER: Poller[Main Poller] Time: 300.0752 s, Poller: 1, Threads: N/A, Devices: 292
2020-10-24 18:20:38 - CMDPHP WARNING: cmd.php poller over ran its polling intervale and therefore ending
This Warning is a good reason to use spine since cmd.php is unable to process any more. So I did changed my mined and start using spine.
Spine Parameters:
Deafault Threads pr Process: 15
Number of PHP Script Servers: 4
Script server timeout Value: 25
Max SNMP OIDs: 10
Because I couldn't see no more Graphs any more I deleted the rra Folder to let cacti rebuild the files. The files are all back again. I restarted poller.php manually and let it run till the end, till it said: OK u:0.28 s:0.43 r:299.35
2020-10-24 18:54:36 - SYSTEM STATS: Time:299.6015 Method:spine Processes:1 Threads:1 Hosts:295 HostsPerProcess:295 DataSources:3304 RRDsProcessed:1355
All my Graphs are desappeared already I can see NO Grpahs any more and I see more cactus pics and all my devices are in unknown state.
This is the spine permission:
-rwsr-sr-x 1 root root 463192 Oct 18 16:20 /usr/local/spine/bin/spine
As I said I use SUSE SLES SP15.1 and my cron is set up to:
because a cacti username is not accepted in suse linux. meaning that if I do chown cactiuser:root to my rra Folder it sais this user name is not recognized. But wwwrun is doing well.
Permissions to rra and my log Folders. drwxr-xr-x 2 wwwrun wwwrun 77824 Oct 24 20:04 rra
lrwxrwxrwx 1 wwwrun www 14 Sep 24 19:21 log -> /var/log/cacti
AM I missing any thing? Can sombody help? Thanks in Advance.
TheWitness wrote: ↑Thu Oct 22, 2020 6:17 pm
When you have problems like this, it's helpful to run spine in debug mode to ensure that the script/snmp is returning the correct data. Like this:
./spine -R -V 3 -f device_id -l device_id -S
Replace "device_id" with your Device ID from the Device Management page. If you are getting good data, then you should look secondarily to maximum values in the RRDfile, and if that fails, look at your Step and Heartbeats and make sure that they align with the Cacti Poller interval.
In other words, if you have your Cron Interval set to 5 minutes, and your poller interval set to 5 minutes, but you are trying to graph at a 1 minute granularity, then Cacti will never work. Much of this is laid out in the documentation.
Hi witness,
I tried the command : /usr/local/spine/bin/spine -R -V 3 -f device_id 22 -C /usr/local/spine/etc/spine.conf all is running well and gave the output value. But in each device, i got this value :
2020/10/26 11:43:36 - SPINE: Poller[1] Device[22] HT[1] DS[520] SS[0] SERVER: /var/www/cacti/scripts/ss_host_cpu.php ss_host_cpu '10.3.100.138' '22' '2:161:500:1:10:public::::::' 'get' 'usage' '4000', output: U
its look like what i see the CPU total is n/a in a graph
any advice on this?
thank you
