problem after migration to cacti0.8.6.c with cactid.0.8.6.d

[p_guerrini]

Hello
I have a redhat 9 on PC with Cacti0.8.6c with with cactid.0.8.6.d.
SInce the migration, cacti does not graph anymore.
I put in the root crontab the following command:
With root: crontab -e
15,30,45,00 * * * * su - cactiuser -c "/usr/bin/php /pkg/cacti/poller.php" > /dev/null 2>&1
In my cacti/log file I have a lot of things browsed and in my rra directory all my files have the new date and hour from the load of cactid.
Thanks do your help

[rony]

Why don't you run it from the users cron?

Code: Select all

crontab -e -u cactiuser

Entry like:

Code: Select all

*/15 * * * * /usr/bin/php /pkg/cacti/poller.php > /dev/null 2>&1 

And do you really want to run it every 15 minutes?

The reason I suggest the above, its that there could be some inherent enviroment issues with using su and paths.

[p_guerrini]

Hello thanks
Can I send you my log and the content of my database. BEcause I spent a lot of time with Fmangeant on the phone and we found npothing except that we can not clear purge the poller cash
Thanks

[rony]

Can you run the poller once from the command line and send me the output?

[p_guerrini]

Hello
I am in week end in France. I will send you the ouput on monday.
Thanks a lot and good week end
Pierre

[p_guerrini]

Hello
I have in my root crontab :
15,35,48,00 * * * * su - cactiuser -c "/usr/bin/php /pkg/cacti/poller.php" > /dev/null 2>&1

All the cacti directory is the cactuiser owner.

see the log in attachement

Redhat 9
[root@ggeuxp37 log]# rpm -qa | grep mysql
mysql-3.23.54a-11
php-mysql-4.2.2-17
mysql-server-3.23.54a-11
mysql-devel-3.23.54a-11

[root@ggeuxp37 log]# rpm -qa | grep php
php-4.2.2-17
php-ldap-4.2.2-17
php-mysql-4.2.2-17
php-imap-4.2.2-17
php-snmp-4.2.2-17

drwxr-xr-x 14 root root 4096 sep 14 2004 rrdtool-1.0.49

Thanks for your help

[p_guerrini]

My last cacti was 8.5.a

[TheWitness]

Run cactid from a command line in debug mode and post output. If you are running mysql 4.1, it is likely a client build issue. You should give the code on the Announcement forum a try if that is the case.

TheWitness

[p_guerrini]

Hello
Sorry my cactid is the 0.8.6e but with the 0.8.6d I have the same pb.
The log of the command cactid is in attachement.
Mysql is :
[root@ggeuxp37 bin]# rpm -qa | grep mysql
mysql-3.23.54a-11
php-mysql-4.2.2-17
mysql-server-3.23.54a-11
mysql-devel-3.23.54a-11
thanks

[TheWitness]

What attachment?

[TheWitness]

Dah, slet that comment. Ok, your problem is too many scripts. Cactid is finishing, but not until after 390 seconds. I need to convert the query_host_partitions to the script server. That will cut 50% off your polling time.

Until then, try to use 2-4 processes to reduce the overall polling cycle.

TheWitness

[TheWitness]

Double slet You are using the query host partitions, you need to migrate to the script server. Closely follow the procedure in the attached link and you will see your polling time decrease significantly.

TheWitness

http://www.cacti.net/downloads/docs/htm ... eries.html

[p_guerrini]

Hello
I did the changement as you said me. It worked since monday but today when I put some new graphs, cacti does not graph.
Thanks to help me

[TheWitness]

Please dump database again and forward via e-mail.

TheWitness

[p_guerrini]

Hello
Sorry for the delay because I was in holidays
See the attachment for the db dump
(mysqldump -l --add-drop-table cacti > mysql.cacti)
Thanks