Strange listing of `discover` job in poller performance

Expected Behavior

In the poller performance page, when I see the performance of the different actions, I should see one line for macsuck, arpnip, discover or other poller type

Current Behavior

But, for an unknown reason, I see the Discover part splited in multiple lines...

Steps to Reproduce (for bugs)

Let the poller run few days
Check the poller performance page

Context

It's worrying when we try to estimate the time of the discover job

Your Environment

Netdisco version used: 2.63.1
SNMP::Info version used: 3.92

Config info (deployment.yml)

#
# NETDISCO 2 CONFIGURATION FILE
#
# Settings in this file override share/config.yml
# See App::Netdisco::Manual::Configuration for more info.

# ------------------
# ESSENTIAL SETTINGS
# ------------------

database:
  name: 'netdiscodb'
  user: 'netdisco'
  pass: 'netdisco'

# -----------------
# LOG SETTINGS
# -----------------
log: 'warning'
log_format: '[%T] {PID: %P} {Level:%L} %m'

# set to false if you MUST maintain backwards compatibility
# with Netdisco 1.x web frontend.
# `````````````````````````````````````````````````````````
safe_password_store: true

# SNMP community string(s)
# ````````````````````````
device_auth:
  - tag: 'snmp-user-ro'
    user: 'snmp-user-ro'
    auth:
      proto: 'SHA'
      pass: 'xxx'
    priv:
      proto: 'AES'
      pass: 'xxx'
  - tag: 'sshbigip'
    driver: cli
    platform: BigIP
    only:
      - 10.208.233.98
      - 10.208.233.99
      - 10.208.236.6
      - 10.208.236.7
    username: netbackup
    ssh_master_opts:
      - "-i"
      - "~/.ssh/rancid_ed25519"

# this is the schedule for automatically keeping netdisco up-to-date
# ``````````````````````````````````````````````````````````````````
schedule:
  discoverall:
    when: '* */6 * * *'
  macwalk:
    when:
      min: '0,30'
  arpwalk:
    when:
      min: '15,45'
  nbtwalk: null
  expire:
    when: '20 23 * * *'
  makerancidconf:
    when: '30 */6 * * *'
  makeansiblehost:
    when: '30 */6 * * *'

# do not discover IP Phones or Wireless Access Points.
# usually these are visible as device neighbors but don't support
# SNMP, which just clogs up the job queue.
# ```````````````````````````````````````````````````````````````
#discover_no_type:
#  - '(?i)phone'
#  - '(?i)cisco\s+AIR-(L|C)AP'
discover_waps: false
discover_phones: false

# number of SNMP workers to run in parallel (in netdisco-backend).
# the default is twice the number of CPU cores. increase this if
# your system has few cores and the schedule is taking too long.
# ```````````````````````````````````````````````````````````````
workers:
  tasks: 'AUTO * 4'
  timeout: 600
  sleep_time: 1
  min_runtime: 0
  max_deferrals: 10
  retry_after: '1 day'

# number of parallel DNS queries for node names
# `````````````````````````````````````````````
dns:
  max_outstanding: 50

# Domain suffix to remove
# `````````````````````````````````````````````
domain_suffix:
  - '.example.com'

# We disable completely the Netbios discover
# `````````````````````````````````````````````````````````````````````
nbtstat_no:
  - 'any'

# Global tuning for MAC suck to take in count users in VLAN
# WARNING: on Cisco switch, you should allow polling on SNMP context for each VLAN
# `````````````````````````````````````````````````````````````````````
macsuck_all_vlans: true

# We clear job that didn't have run after 2 days
# `````````````````````````````````````````````````````````````````````
expire_jobs: 2

# Turn this on to have Netdisco do a reverse lookup of the device sysName.0 field 
# to use as the management IP address for a device.
# `````````````````````````````````````````````````````````````````````
reverse_sysname: true

# Web tune parameters
# `````````````````````````````````````````````````````````````````````
table_pagesize: 50

# LDAP authentication
# `````````````````````````````````````````````````````````````````````
ldap:
  servers:
    - 'ad.example.com'
  user_string: '%USER%@ad.example.com'
  opts:
    debug: 0

# Discover limit
#
discover_only:
  - '10.0.0.0/8'
  - '100.64.0.0/10'
  - '192.168.0.0/16'
discover_no:
  - 'excluded.example.com'

# vim: tabstop=2 expandtab

netdisco / netdisco

Strange listing of `discover` job in poller performance #1079