Comments on: MySQL Cacti templates version 1.1.1 released http://www.xaprb.com/blog/2008/10/15/mysql-cacti-templates-version-111-released/ Stay curious! Thu, 09 Feb 2012 09:56:43 +0000 hourly 1 http://wordpress.org/?v=3.2.1 By: Xaprb http://www.xaprb.com/blog/2008/10/15/mysql-cacti-templates-version-111-released/#comment-15537 Xaprb Thu, 11 Dec 2008 15:17:31 +0000 http://www.xaprb.com/blog/?p=626#comment-15537 John, please submit a bug on the Google Code project. The templates should import fine on newer versions if you generate them for an older version, but if that isn't the case, submit a bug for that too. John, please submit a bug on the Google Code project. The templates should import fine on newer versions if you generate them for an older version, but if that isn’t the case, submit a bug for that too.

]]> By: John http://www.xaprb.com/blog/2008/10/15/mysql-cacti-templates-version-111-released/#comment-15529 John Wed, 10 Dec 2008 21:34:19 +0000 http://www.xaprb.com/blog/?p=626#comment-15529 Trying to create custom template and I am getting a XML Parse error. I am using version 0.8.7b and when i choose --cactiver it does not allow me to use 0.8.7b. It only shows and allows up to 0.8.6j. Any suggestions? Trying to create custom template and I am getting a XML Parse error. I am using version 0.8.7b and when i choose –cactiver it does not allow me to use 0.8.7b. It only shows and allows up to 0.8.6j. Any suggestions?

]]> By: Mikael Fridh http://www.xaprb.com/blog/2008/10/15/mysql-cacti-templates-version-111-released/#comment-15390 Mikael Fridh Thu, 13 Nov 2008 17:24:22 +0000 http://www.xaprb.com/blog/?p=626#comment-15390 Yeah but the poller still executes ss_get_mysql_stats.php 24 times. 100 database servers with my old templates = 100 script executions. 100 database servers with MySQL Cacti templates = 2400 script executions. Indeed, Cacti's design is not optimal, and the lack of a decent polling/caching API is plagueing it. I still think getting people to increase their fgets() sizes and doing all the polling in one call is optimal :). It just sucks that when you add graphs etc, you sometime get duplicated data sources. *grr* I can perhaps commit to looking at some of the code december 1, after my vacation. Thanks for your great work so far :) Yeah but the poller still executes ss_get_mysql_stats.php 24 times.

100 database servers with my old templates = 100 script executions.
100 database servers with MySQL Cacti templates = 2400 script executions.

Indeed, Cacti’s design is not optimal, and the lack of a decent polling/caching API is plagueing it.

I still think getting people to increase their fgets() sizes and doing all the polling in one call is optimal :). It just sucks that when you add graphs etc, you sometime get duplicated data sources. *grr*

I can perhaps commit to looking at some of the code december 1, after my vacation. Thanks for your great work so far :)

]]> By: Xaprb http://www.xaprb.com/blog/2008/10/15/mysql-cacti-templates-version-111-released/#comment-15383 Xaprb Wed, 12 Nov 2008 14:13:00 +0000 http://www.xaprb.com/blog/?p=626#comment-15383 Mikael, it would be great to have another commiter on the code project, so if you want SVN access just ask. Mikael, it would be great to have another commiter on the code project, so if you want SVN access just ask.

]]> By: Xaprb http://www.xaprb.com/blog/2008/10/15/mysql-cacti-templates-version-111-released/#comment-15382 Xaprb Wed, 12 Nov 2008 14:10:58 +0000 http://www.xaprb.com/blog/?p=626#comment-15382 Mikael, my script should execute once per polling cycle per host, that is one of the design elements I spent a lot of time on. Are you sure the caching is working right? If not please put a bug report on the Google Code project. The point of my design is to work around Cacti's limitations here: http://forums.cacti.net/viewtopic.php?t=26448#130332 Cacti's design forces a tradeoff among various complaints, and caching results is the only good way around it right now. As for Handler_* stats, I have no clue why I omitted those! Please submit a bug report on that. Mikael, my script should execute once per polling cycle per host, that is one of the design elements I spent a lot of time on. Are you sure the caching is working right? If not please put a bug report on the Google Code project. The point of my design is to work around Cacti’s limitations here: http://forums.cacti.net/viewtopic.php?t=26448#130332 Cacti’s design forces a tradeoff among various complaints, and caching results is the only good way around it right now.

As for Handler_* stats, I have no clue why I omitted those! Please submit a bug report on that.

]]>