InfluxDB/Grafana/Bosun + Cacti = CereusTransport - Updated!

[lucajet]

Could be possible to pass to influxdb just a subset of the data sources instead of everything?

[phalek]

Hi

Currently the plugin doesn't allow this, it just graps everything and sends it off to influxdb

[bigblue]

Hi,

I have a problem getting the plugin to work:
cacti logs:
"ERROR: A DB Exec Failed!, Error: '1114', SQL: "INSERT INTO `plugin_CereusTransporter_data
"
log mysql:
"mysqld: The table 'plugin_CereusTransporter_data' is full
"

I tried to increase the @@max_heap_table_size to 1Gb, but nothing worked.
The influxdb database does not receive any data.

Can you help me?

Cacti version: 0.8.8h
CereusTransporter version: 0.64
Devices > 1200 Hosts
Graph Management > 70K Graphs

Regards

[phalek]

Drop the table and re-create it:

Code: Select all

mysql> drop table plugin_CereusTransporter_data;

mysql> CREATE TABLE `plugin_CereusTransporter_data` (`timestamp` VARCHAR(1024) NOT NULL DEFAULT '0',`local_data_id` INT(11) NOT NULL DEFAULT '0',`key` VARCHAR(1024) NOT NULL DEFAULT '0',`value` VARCHAR(1024) NOT NULL DEFAULT '0',INDEX `local_data_id` (`local_data_id`),INDEX `key` (`key`)) COMMENT='NMID CereusTransporter Data' COLLATE='latin1_swedish_ci' ENGINE=MyISAM;

[bigblue]

Drop the table and re-create it:

Code: Select all

mysql> drop table plugin_CereusTransporter_data;

mysql> CREATE TABLE `plugin_CereusTransporter_data` (`timestamp` VARCHAR(1024) NOT NULL DEFAULT '0',`local_data_id` INT(11) NOT NULL DEFAULT '0',`key` VARCHAR(1024) NOT NULL DEFAULT '0',`value` VARCHAR(1024) NOT NULL DEFAULT '0',INDEX `local_data_id` (`local_data_id`),INDEX `key` (`key`)) COMMENT='NMID CereusTransporter Data' COLLATE='latin1_swedish_ci' ENGINE=MyISAM;

great!
Thank you, I have no more errors in cacti log and mysqldb.
Cacti log:
"01/16/2020 01:28:34 PM - CereusTransporter: Poller [0] INFO: Adding data to influxdb"

The plugin_CereusTransporter_data table is well filled, but my InfluxDB database is not receiving any packets:

# tcpdump -i enp0s3 port 8086
tcpdump: verbose output suppressed, use -v or -vv for full protocol decode
listening on enp0s3, link-type EN10MB (Ethernet), capture size 262144 bytes
^C
0 packets captured
2 packets received by filter
0 packets dropped by kernel

# influx -precision rfc3339
Connected to http://localhost:8086 version 1.7.9
InfluxDB shell version: 1.7.9
> SHOW DATABASES
name: databases
name
----
_internal
> quit

However, I configured the plugin like that:
influxdb://xxx.xxx.185.222:8086/cacti

[phalek]

you're using localhost but referr to a different IP for the URL

Can you open the Influx Webpage using the IP ? It may just be a firewall problem.

[bigblue]

Sorry... I didn't understand influxdb:// means local connection...

I configured the plugin in http, but still nothing in influxdb.

Is there a special configuration to do for influxdb?
Is it necessary to create the DB manually?

[phalek]

Where is influxdb running, on the same host as cacti ( then use influxdb://localhost ) or on a different server ( then use influxdb://<remote ip> ... and make sure the firewall actually allows the connection to the influxdb port ) ?

Also if you're using PHP 7 then this will probably not work

[phalek]

Oh, and i didn't see the cacti database in influxdb. did you create it at all ?

[bigblue]

Yes, I'm using another server, so I changed it to "influxdb://<remote ip>:8086/cacti".
The firerwall is disabled.
The telnet <remote ip> 8086 works fine.

But still nothing comes to influxdb.

My php version:
$ php -v
PHP 5.4.16 (cli) (built: Nov 1 2019 16:04:20)

I didn't create a table in influxdb, now I have.
Is there anything else we can do?

[bigblue]

I did some php debugging, the program doesn't pass the next command:
$polling_data = db_fetch_assoc( "SELECT `timestamp`, `local_data_id`, `key`, `value` FROM plugin_CereusTransporter_data ORDER BY `timestamp`,`local_data_id`,`key` );

Do you know why?

[phalek]

What error message do you get when calling this SQL statement from within mysql ?

[bigblue]

There isn't one... it runs and apparently finishes normally:
MariaDB [cacti]> SELECT `timestamp`, `local_data_id`, `key`, `value` FROM plugin_CereusTransporter_data ORDER BY `timestamp`,`local_data_id`,`key`
-> ;
+------------+---------------+---------------------+------------------+
| timestamp | local_data_id | key | value |
+------------+---------------+---------------------+------------------+
| 1579269002 | 19093 | 5min_cpu | 22 |
| 1579269002 | 19094 | cisco_memfree | 4737352 |
| 1579269002 | 19095 | cisco_memused | 17571344 |
| 1579269002 | 23640 | 5min_cpu | 23 |
...
...
| 1579269040 | 1330015 | 5min_cpu | 8 |
| 1579269040 | 1330016 | cisco_memfree | 25587620 |
| 1579269040 | 1330017 | cisco_memused | 11599784 |
| 1579269040 | 1330018 | arp_available | 5936 |
| 1579269040 | 1330018 | arp_used | 1434 |
+------------+---------------+---------------------+------------------+
209110 rows in set (3.72 sec)

[bigblue]

Hi,

I'm trying to automate the creation of dashborad in grafana.
Is there anything (plguins) at all done?
If not, what is the best (simplest) way to do it?

Thanks
BB

[phalek]

It looks like there's a quote missing to the end of that SQL statement.

Code: Select all

... `key` ")