THOLD broken-ness after upgrade/migration

[GlueGuy]

I created a new server running CentOS 6.7 to host an updated version of cacti (0.8.8f).

The "old" cacti server is CentOS 5.11 running 0.8.8a cacti, along with several plugins, including THOLD.

The new system is running base cacti 0.8.8f just fine. In fact, it is monitoring ~~ 1,000 SNMP devices very well.

HOWEVER, when I imported all the SQL from the old system onto the new system, there were tables/entries that made THOLD inop.

So I disabled/de-installed THOLD, and moved the plugin to a different directory.

When I do an SQL dump, there are still tables & bits from THOLD in the database. Is there some simple way to purge these leftovers (which I think is the source of the problem)?

[GlueGuy]

After "disabling/removing" the THOLD plugin, there were still a couple of tables (e.g. plugin_thold_host_failed), plus a BUNCH of entries in the settings table.

I removed all the tables that had "thold" in their name, and I removed all the rows in the settings table that mentioned "thold".

I then placed the thold folder back into the plugins folder, and reinstalled/enabled.

It works now.

I do have to re-enter all the thresholds [about 30 of them ], but at least it's working.