Host seen "Down" in cacti whereas UP

[stekut]

Hi all,

here is my problem :

we have to rename all servers, so I wrote a script to do it in cacti DB.
Let's say in cacti my host is A.
Everything is fine, PING and SNMP works (Graphs are OK)

I launched my script, now hostname is B. PING still works, SNMP too (when I go in the device, I see the SNMP information).

But now in my device list, I see that my host (now named B) is seen "Down" by cacti...it's a real issue for me since Cacti stop graphing ! What's wrong ?

My "Downed device detection" is set as "SNMP", so it should be ok.
I still have the SNMP Information working and Cacti still see my interfaces when I reload my "SNMP - Interface statistics" data query.

Does somebody have an idea ?

Thanks for your help.

[TheWitness]

If you are using spine, you can poll at level 3 in read only mode and see any errors.

Code: Select all

./spine -V 3 -f <host_id> -l <host_id> -R

TheWitness

[stekut]

It would be too easy...we don't use spine

In fact, I found the solution.
When I modified the name through MySQL, I updated the "host" table to modify hostname and description, the "graph_templates_graph" table to update title_cache and that's all.

One more table needed to be updated : "poller_item" which store hostname too !

Now it's OK.

Anyway, thanks for your help.