thold plugin 0.4.6 screenshots

[skatter]

Please disable STRICT for the SQL_MODE setting in MySQL (/etc/my.cnf) and then restart MySQL. It looks like quite a bit failed with the default, which Cacti is not compatible with. After which, you should consider starting over if you are not too far along.

Well, trying to dig into my.cnf is basically showing that there isn't one configured. I installed MySQL from the FreeBSD ports tree, and the only .cnf files are in /usr/share/mysql/my-[small|medium|large|huge].cnf; the start scripts indicates that my.cnf will be called from /var/db/mysql:

Code: Select all

command_args="-c -f /usr/local/bin/mysqld_safe --defaults-extra-file=${mysql_dbdir}/my.cnf --user=${mysql_user} --datadir=${mysql_dbdir} --pid-file=${pidfile} ${mysql_args}"

but there is no my.cnf in /var/db/mysql.

Also, in the existing my-*.cnf files, I don't see SQL_MODE specified at all, let alone as STRICT. Can I just copy one of the my-size.cnf files to /var/db/mysql, restart MySQL, and see if that fixes it?

[TheWitness]

my.cnf is in /etc/

Code: Select all

mysql> show global variables like '%mode%';
+--------------------------+--------+
| Variable_name            | Value  |
+--------------------------+--------+
| innodb_autoinc_lock_mode | 1      |
| slave_exec_mode          | STRICT |
| sql_mode                 |        |
+--------------------------+--------+
3 rows in set (0.01 sec)

The key word is 'sql_mode'.

TheWitness

[skatter]

Here's the output from my instance of MySQL:

Code: Select all

mysql> show global variables like '%mode%';
+--------------------------+--------+
| Variable_name            | Value  |
+--------------------------+--------+
| innodb_autoinc_lock_mode | 1      |
| innodb_strict_mode       | OFF    |
| slave_exec_mode          | STRICT |
| sql_mode                 |        |
+--------------------------+--------+
4 rows in set (0.00 sec)

[TheWitness]

Not sure what your problem is then. Odd.

[skatter]

Not sure what your problem is then. Odd.

Odd indeed, which is why I tried to manually import the thold.sql file to get the tables created, so that I could subsequently use thold.

Do you have any suggestions on steps I can take to identify the problem here? From the GUI thold appears to install normally, but the underlying tables are lacking.

[TheWitness]

Post the output from your apache error log and from the Cacti log immediately upon install.

[uhtred]

I found this, but only for 'Time Based' alarms. Can you confirm in svn1910.

TheWitness

I did try svn1910 & 1911, but my cacti crash, if i mv thold_functions.php into thold folder!
so i have to go back to svn1906.

[TheWitness]

It's not thold_functions.php causing that. THold 0.4.8-1 is released.

[skatter]

Post the output from your apache error log and from the Cacti log immediately upon install.

Here's my apache error log:

Code: Select all

[Fri Oct 21 13:31:24 2011] [error] [client 1.2.3.4] PHP Notice:  Undefined index: keys in /usr/local/share/cacti/lib/plugins.php on line 174, referer: http://hostname/cacti/plugins.php
[Fri Oct 21 13:31:24 2011] [error] [client 1.2.3.4] PHP Notice:  Undefined index: keys in /usr/local/share/cacti/lib/plugins.php on line 174, referer: http://hostname/cacti/plugins.php

I didn't see anything in cacti's log (set to HIGH) during the install/enabling of thold, but I did get some of the following after the poller ran:

Code: Select all

10/21/2011 01:34:07 PM - CMDPHP: Poller[0] ERROR: SQL Assoc Failed!, Error:'1146', SQL:"SELECT * FROM thold_data WHERE thold_enabled='on' AND tcheck=1"
10/21/2011 01:33:08 PM - CMDPHP: Poller[0] ERROR: A DB Exec Failed!, Error:'1146', SQL:"UPDATE thold_data SET tcheck=0'
10/21/2011 01:33:08 PM - CMDPHP: Poller[0] ERROR: A DB Exec Failed!, Error:'1146', SQL:"DELETE FROM plugin_thold_log WHERE time<1316550788'

I also see this error in my apache error log when I click on the Thresholds link:

Code: Select all

[Fri Oct 21 13:38:22 2011] [error] [client 1.2.3.4] PHP Warning:  Invalid argument supplied for foreach() in /usr/local/share/cacti/plugins/thold/listthold.php on line 271, referer: http://hostname/cacti/plugins.php

edit: I upgraded to 0.4.8 with the same results.

[TheWitness]

What's your mysql version and PIA version. I think that this will only be compatible with PIA 2.9/3.0.

TheWitness

[skatter]

What's your mysql version and PIA version. I think that this will only be compatible with PIA 2.9/3.0.

TheWitness

Cacti Version: 0.8.7g, Plugin Architecture Version: 2.9
mysql-client-5.1.49 Multithreaded SQL database (client)
mysql-server-5.1.41 Multithreaded SQL database (server)

[TheWitness]

Code: Select all

delete from settings where name='plugin_thold_version'

[TheWitness]

Then uninstall and reinstall

[skatter]

Code: Select all

delete from settings where name='plugin_thold_version'

Code: Select all

mysql> delete from settings where name='plugin_thold_version';
Query OK, 1 row affected (0.01 sec)

Then I disabled/uninstalled -> installed/enabled thold from PIA and I'm still seeing the same error in my apache error log, and I still don't see any new thold related tables in my cacti DB.

[TheWitness]

Update to 0.4.8-2. There was a conflict with weathermap.