[SOLVED] No graph after migration 0.8.6j to 0.8.7a

[phi00611]

Hello,

I have a serious problem. I migrated version 0.8.6j of cacti to 0.8.7a. But since my migration graphs remain empty.

The log file cacti.log is empty.

Could you help me?

Thanks.

[phi00611]

I rebooted the poller, and now I have my data in graphs.
But the graphs on the network interfaces are still empty.

[BSOD2600]

Follow these steps http://docs.cacti.net/node/283

[phi00611]

Hello,

I ran the "cmd.php 73 73" command manualy in debug mode level. And I get the following data in the log file :

01/14/2008 09:23:22 AM - POLLER: Poller[0] NOTE: There are no items in your poller for this polling cycle!
01/14/2008 09:23:22 AM - CMDPHP: Poller[0] DEBUG: SQL Exec: "UPDATE poller_item SET rrd_next_step = rrd_step - 300 WHERE (rrd_next_step < 0 AND host_id >= 73 AND host_id <= 73)"
01/14/2008 09:23:22 AM - CMDPHP: Poller[0] DEBUG: SQL Exec: "UPDATE poller_item SET rrd_next_step = rrd_next_step - 300 WHERE (host_id >= 73 AND host_id <= 73)"
01/14/2008 09:23:22 AM - CMDPHP: Poller[0] DEBUG: SQL Cell: "SELECT count(*) FROM poller_item WHERE (action=2 AND host_id >= 73 AND host_id <= 73 AND rrd_next_step <= 0)"
01/14/2008 09:23:22 AM - CMDPHP: Poller[0] DEBUG: SQL Assoc: "SELECT * FROM poller_item WHERE (host_id >= 73 AND host_id <= 73 AND rrd_next_step <= 0) ORDER by host_id"
01/14/2008 09:23:22 AM - CMDPHP: Poller[0] DEBUG: SQL Assoc: " SELECT * FROM host WHERE (disabled = '' AND id >= 73 AND id <= 73) ORDER by id"
01/14/2008 09:23:21 AM - CMDPHP: Poller[0] DEBUG: SQL Assoc: "select realm_id from user_auth_realm where user_id=1"
01/14/2008 09:23:21 AM - CMDPHP: Poller[0] DEBUG: SQL Cell: "select username from user_auth where id=1"
01/14/2008 09:23:21 AM - CMDPHP: Poller[0] DEBUG: SQL Assoc: "select user_auth_realm.realm_id from user_auth_realm where user_auth_realm.user_id='1' and user_auth_realm.realm_id='15'"
01/14/2008 09:23:21 AM - CMDPHP: Poller[0] DEBUG: SQL Cell: "select id from user_auth where username='0'"
01/14/2008 09:23:21 AM - CMDPHP: Poller[0] DEBUG: SQL Cell: "select cacti from version"
01/14/2008 09:23:19 AM - CMDPHP: Poller[0] DEBUG: SQL Exec: "insert into poller_time (poller_id, start_time, end_time) values (0, NOW(), NOW())"
01/14/2008 09:23:19 AM - CMDPHP: Poller[0] Time: 17.8285 s, Theads: N/A, Hosts: 72
01/14/2008 09:23:19 AM - PHPSVR: Poller[0] DEBUG: PHP Script Server Shutdown request received, exiting

I don' interpret this results.
Can you help me?

[phi00611]

I inventoried all my hosts (total 73 hosts), and on 5 hosts I have data in network graphs, but on all over the network graphs are empty.

All my servers are in the same network, and I get the data when using the command snmpwalk.

[gandalf]

Try

Code: Select all

php -q poller.php --force --debug

Reinhard

[phi00611]

Hello,

I used the command

Code: Select all

php -q poller.php --force --debug

, and I obtain the following result :

All tests are OK, and the final status is :

"AM - SYSTEM STATS: Time:17.6159 Method:cmd.php Processes:1 Threads:N/A Hosts:73 HostsPerProcess:73 DataSources:1129 RRDsProcessed:722"

I have patch the cacti version with 2 officials patches, and I rebuild the poller cache.

And the network graphes are always empty.

[phi00611]

The servers have no graphs are bonding configured.

If I add graphs for eth0 and eth1 on these servers, I get always empty graph with the value "nan".

However, the network graphs on servers not having bonding interface functioning normally.

[phi00611]

Hello,

I checked my configuration, and I noticed that there was something odd in the "data template" "interface-traffic".
In the parameter "Associated RRA's", I have 2 lines "Hourly (1 minute Average)" including one selected. [/img]

I have deselected "Hourly (1 minute Average)" and I rebuilded poller cache, and now I have data in my network graphes. But only traffic outbound is displayed.

Traffic inbound have always "nan" value.

[gandalf]

Please post the table under "Data Sources -> RRAs"
Definitively, there should be only one line "Hourly ..."
Reinhard

[phi00611]

Below is the table "data sources>RRA's" :
[/img]

Thanks.

[fmangeant]

When importing the new Apache templates, did you check the "Use custom RRA settings from the template" option ?

[phi00611]

I have not yet installed the plugin apacheStat on this server.

[gandalf]

Below is the table "data sources>RRA's" :
[/img]

Thanks.

This is definitively broken. Either you manually added the second hourly rra or you've got it by importing a template, as fmangeant suggested.
You may investigate and search all your data templates to find a place were the second entry is used. To ease searching, I recommend renaming one of them so you will find it easily on the data templates. When editing the rra entries, you will find their id listed in the URL string. In general, the higher id is causing the trouble. So please rename this one
Reinhard

[phi00611]

Hello,

I renamed in "data source> RRA's" option "Hourly." I deleted all my network graphs and I recreated it.

Now, I have the data in graphs network.

Thanks for your help.