[Bug]: many chart titles contain instance-specific information which is "wrong"

[ilyam8]

Bug description

It is not explicitly mentioned in our documentation, but it appears that having instance-specific information in a chart title should be avoided because it causes some problems on the Cloud.

The biggest offenders:

• diskspace.plugin

  Disk Space Usage for X
  Disk Files (inodes) Usage for X

• Anomaly detection

  Anomaly detection dimensions for host X
  Anomaly detector window length for host X
  Anomaly events triggered for host X
  Percentage of anomalous dimensions for host X

• ML

  Training step statistics for host X
  Training thread CPU usage for host X
  Prediction thread CPU usage for host X
  Time it took to run prediction for host X

• go.d.plugin

  Execution Time for X

• python.d.plugin

  Execution Time for X

• python.d.plugin (sensors)

  X temperature

• charts.d.plugin (sensors)

  X Temperature

• proc.plugin (btrfs)

  BTRFS System Allocation for X
  BTRFS Metadata Allocation for X
  BTRFS Physical Disk Allocation for X
  BTRFS Data Allocation for X

• proc.plugin (softnet)

  CPUX softnet_stat

• proc.plugin (sofirqs)

  CPUX softirqs

• proc.plugin (interrupts)

  CPUX Interrupts

• fping.plugin (should make a PR to the upstream repo)

  FPing Quality for host X
  FPing Packets for host X
  FPing Latency for host X

• Netdata webserver

  Netdata web server thread No X CPU usage

Expected behavior

No instance-specific information in chart titles.

Steps to reproduce

n/a

Installation method

kickstart.sh

System info

n/a

Netdata build info

Version: netdata v1.33.1-306-gf896148a3

Additional info

No response

[ilyam8]

@vlvkobal we need to fix https://github.com/netdata/ioping