[Netdisco] Strange backup job errors

[Ben C. <ben...@gm...>]

I started getting errors from cron a few days ago - no changes were made on
the server, the command is just netdisco -B -b

DBD::Pg::st execute failed: ERROR:  relation "node" does not exist
DBD::Pg::st fetchrow_hashref failed: no statement executing
DBD::Pg::st execute failed: ERROR:  relation "node_ip" does not exist
DBD::Pg::st fetchrow_hashref failed: no statement executing
DBD::Pg::st execute failed: ERROR:  relation "node_ip" does not exist
DBD::Pg::st fetchrow_hashref failed: no statement executing
DBD::Pg::st execute failed: ERROR:  relation "node_nbt" does not exist
DBD::Pg::st fetchrow_hashref failed: no statement executing
DBD::Pg::st execute failed: ERROR:  relation "node_nbt" does not exist
DBD::Pg::st fetchrow_hashref failed: no statement executing


These messages come about 5 hours after the job should have run. They don't
make much sense as the relations do exist. Running the job manually in debug
it just hangs on the last line..

n e t  d i s c o
--------------------------------------------------
Using Config File : /usr/local/netdisco/netdisco.conf
nightly() - Starting nightly cleanup and backup routines
  Cleaning and clearing data...
  expire_devices: Deleting Devices that are older than 180 days old.

expire_data(device,180 days,archive_only:0)

  expire_nodes_archive: Deleting Archived Node data older than 180 days old.

expire_data(node,180 days,archive_only:1)
Deleted 0E0 rows from node.
Deleted 0E0 rows from node_ip.
Deleted 0E0 rows from node_nbt.
  expire_nodes: Deleting Nodes older than 180 days old.

expire_data(node,180 days,archive_only:0)
Deleted 0E0 rows from node.
Deleted 0E0 rows from node_ip.
Deleted 0E0 rows from node_nbt.
  expire_processes: Deleting processes created over 1 day(s) ago.

expire_data(process,1 days,archive_only:0)
Deleted 0E0 rows from process table.
Database Cleanup :
load_old_devices()
  Deleting nodes that are actually device ports...
    0E0 MAC entries deleted
    156 ARP entries deleted
  Checking for Nodes that exist on nonexistent ports:
  Removing nodes that are listed on uplink ports...
10.10.0.81 / ge.1.8 -> 192.168.26.1 (not deleting, not discovered)
10.10.0.171 / ge.1.4 -> 150.203.153.204 (not deleting, not discovered)
10.10.2.51 / ge.3.14 -> 150.203.62.204 (not deleting, not discovered)
10.10.8.51 / ge.2.33 -> 203.10.57.252 (not deleting, not discovered)
10.10.10.90 / ge.2.33 -> 203.10.57.252 (not deleting, not discovered)
10.10.16.8 / ge.1.35 -> 130.56.58.11 (not deleting, not discovered)
10.10.16.9 / ge.1.5 -> 130.56.68.20 (not deleting, not discovered)
10.10.16.31 / ge.1.46 -> 130.56.58.12 (not deleting, not discovered)
10.10.18.4 / tg.4.203 -> 172.20.1.48 (not deleting, not discovered)
10.10.18.4 / tg.5.203 -> 172.20.1.47 (not deleting, not discovered)
10.10.18.5 / ge.1.11 -> 192.168.1.20 (not deleting, not discovered)
150.203.202.96 / GigabitEthernet2/8 -> 150.203.4.18 (not deleting, not
discovered)
Done.
expire_ips()

Any ideas where to look?

cheers,

Ben