Boost 2.4 - Released

[TheWitness]

All,

Just a few minor changes here.

Code: Select all

-bug: [Boost 2.3] Include line referring to thold plugin
-bug: Should be db_fetch_row and not db_fetch_cell
-feature: Allow specification of memory limits

TheWitness

[Atkatla]

Thanks for your great work, witness.

How is the upgrade path? I am currently running cacti 0.8.7b with spine 0.8.7a and boost 1.7.
I want to upgrade cacti to 0.8.7d and boost too. Can I go directly from 1.7 to 2.4?

Atk.

[TheWitness]

If running pia 2.x, simply copy the new files in place and goto the plugin management page. Irs automatic. Then, if you were previous forcing boost through cron, remove that entry. Windows is a bit more complicated.

TheWitness

[finn]

Hey.

I notice the patches and latest boost only has support for structured paths in the actual setup script. Meaning that if new hosts are added and you are using boost_server, the directories will still not be created?

I am runing 087d with boost 241 and had to disable boost_server and have the updates done via the http launched process in order to get new host rrd files created.

I am starting to get the gist that web updates are the prefered method of operation when running boost rather than boost_server?

[TheWitness]

Hey.

I notice the patches and latest boost only has support for structured paths in the actual setup script. Meaning that if new hosts are added and you are using boost_server, the directories will still not be created?

I am runing 087d with boost 241 and had to disable boost_server and have the updates done via the http launched process in order to get new host rrd files created.

I am starting to get the gist that web updates are the prefered method of operation when running boost rather than boost_server?

Yes, I see this as a problem in the current code with Structured pathes enabled. Can you please open a bug report at http://bugs.cacti.net

Thanks,

TheWitness

[TheWitness]

Hey.

I notice the patches and latest boost only has support for structured paths in the actual setup script. Meaning that if new hosts are added and you are using boost_server, the directories will still not be created?

I am runing 087d with boost 241 and had to disable boost_server and have the updates done via the http launched process in order to get new host rrd files created.

I am starting to get the gist that web updates are the prefered method of operation when running boost rather than boost_server?

I have made a patch. I can PM you the modified file. Simply PM me first.

TheWitness

[finn]

done:

http://bugs.cacti.net/view.php?id=1427

Hey.

I notice the patches and latest boost only has support for structured paths in the actual setup script. Meaning that if new hosts are added and you are using boost_server, the directories will still not be created?

I am runing 087d with boost 241 and had to disable boost_server and have the updates done via the http launched process in order to get new host rrd files created.

I am starting to get the gist that web updates are the prefered method of operation when running boost rather than boost_server?

Yes, I see this as a problem in the current code with Structured pathes enabled. Can you please open a bug report at http://bugs.cacti.net

Thanks,

TheWitness

[finn]

I have made a patch. I can PM you the modified file. Simply PM me first.

TheWitness

That's fine. I will wait for the next update or whatever. Things are running fine since disabling boost_server so I have no immediate need to go back to using boost_server. Cheers though

[Mikf]

hello
why in boost_sql_myisam.sql table poller_output_boost_processes
is type ENGINE=MEMORY and not ENGINE=MyISAM ?

Jerome.

[TheWitness]

Cause it's better that way. Not all tables need to be myisam as the name suggests.

TheWitness

[dononeil]

Is there any documentation available on how to interact with the RRD's through the boost server? We'd like to 'extract' data from the RRD's for additional analysis on another machine and I thought the boost server would be the perfect way to do this.

[dononeil]

We just rebooted our server, and since our boost tables are in RAM we lost everything that was collected since the last flush to the RRD files...

How can I force a flush to the RRD files before I do a reboot to avoid the data loss in the future?

[TheWitness]

Two ways:

Code: Select all

php -q /var/www/html/cacti/plugins/boost/poller_boost.php --force

or:

Code: Select all

mysql -e "alter table poller_output_boost, engine=myisam;"

then when you reboot:

Code: Select all

mysql -e "alter table poller_output_boost, engine=memory;"

If the table is large, you might want to both in the above order.

TheWitness

[bldewolf]

Hey, I upgraded boost recently and noticed this fix still hasn't been applied. This means that poller_boost.php, if run when it's already running, will assume the previous run has crashed and start running itself. This will result in multiple poller_boost.phps running at the same time.

Code: Select all

diff -ur a/poller_boost.php b/poller_boost.php
--- a/poller_boost.php 2009-02-22 09:29:22.000000000 -0800
+++ b/poller_boost.php      2009-06-08 13:19:47.000000000 -0700
@@ -64,7 +64,7 @@

        /* detect a process that has overrun it's warning time */
        if (substr_count($boost_poller_status, "running")) {
-               $status_array = explode(":", $boost_poller_status);
+               $status_array = explode("time:", $boost_poller_status);

                if (!empty($status_array[1])) {
                        $previous_start_time = strtotime($status_array[1]);

[timi]

hi i have a problem, with boost 2.4 and cacti 0.8.7d

Code: Select all

06/16/2009 04:45:10 PM - CMDPHP: Poller[0] ERROR: A DB Exec Failed!, Error:'1114', SQL:"INSERT INTO poller_output_boost (local_data_id, rrd_name, time, output) VALUES ('8948','traffic_out','2009-06-16 16:45:08','2897437436'), ('9257','errors_out','2009-06-16 16:45:09','0'), ('9257','discards_in','2009-06-16 16:45:09','0'), ('9256','traffic_out','2009-06-16 16:45:09','0'), ('9256','traffic_in','2009-06-16 16:45:09','0'), ('9255','nonunicast_in','2009-06-16 16:45:09','0'), ('9255','nonunicast_out','2009-06-16 16:45:09','0'), ('9254','unicast_out','2009-06-16 16:45:09','0'), ('9254','unicast_in','2009-06-16 16:45:09','0'), ('9253','errors_out','2009-06-16 16:45:09','0'), ('9253','discards_out','2009-06-16 16:45:09','0'), ('9253','discards_in','2009-06-16 16:45:09','0'), ('9253','errors_in','2009-06-16 16:45:09','0'), ('9252','traffic_out','2009-06-16 16:45:09','0'), ('9252','traffic_in','2009-06-16 16:45:09','0'), ('9251','nonunicast_in','2009-06-16 16:45:09','0'), ('9251','nonunicast_out','2009-06-16 16:45:09','0'), ('9250','unicast_out','2009-06-16 16:45:09','0'), ('9250','unicast_in','2009-06-16 16:45:09','0'), ('9249','errors_out','2009-06-16 16:45:09','0'), ('9249','discards_out','2009-06-16 16:45:09','0'), ('9249','discards_in','2009-06-16 16:45:09','0'), ('9249','errors_in','2009-06-16 16:45:09','0'), ('9248','traffic_out','2009-06-16 16:45:09','0'), ('9248','traffic_in','2009-06-16 16:45:09','0'), ('9247','nonunicast_in','2009-06-16 16:45:09','0'), ('9247','nonunicast_out','2009-06-16 16:45:09','0'), ('9246','unicast_out','2009-06-16 16:45:09','0'), ('9246','unicast_in','2009-06-16 16:45:09','0'), ('9245','errors_out','2009-06-16 16:45:09','0'), ('9245','discards_out','2009-06-16 16:45:09','0'), ('9245','discards_in','2009-06-16 16:45:09','0'), ('9245','errors_in','2009-06-16 16:45:09','0'), ('9244','traffic_out','2009-06-16 16:45:09','0'), ('9244','traffic_in','2009-06-16 16:45:09','0'), ('9243','nonunicast_in','2009-06-16 16:45:09','12'), ('9243','nonunicast_out','2009-06-16 16:45:09','417640'), ('9242','unicast_out','2009-06-16 16:45:09','2353'), ('9242','unicast_in','2009-06-16 16:45:09','352'), ('9241','errors_out','2009-06-16 16:45:09','0'), ('9241','discards_out','2009-06-16 16:45:09','0'), ('9241','discards_in','2009-06-16 16:45:09','0'), ('9241','errors_in','2009-06-16 16:45:09','0'), ('9240','traffic_out','2009-06-16 16:45:09','32573342'), ('9240','traffic_in','2009-06-16 16:45:09','34735'), ('9239','nonunicast_in','2009-06-16 16:45:09','0'), ('9239','nonunicast_out','2009-06-16 16:45:09','0'), ('9238','unicast_out','2009-06-16 16:45:09','0'), ('9238','unicast_in','2009-06-16 16:45:09','0'), ('9237','errors_out','2009-06-16 16:45:09','0') ON DUPLICATE KEY UPDATE output=VALUES(output)'

what can it be ?