v1.8.0

New Features

• Add data-dir support for TiCDC in tiup-playground (#1631, @nexustar)
• Add support of using custom files as input of edit-config, and support dumping the current full config to a file with show-config command in tiup-cluster (#1637, @haiboumich)
• Add support of next-generation monitor (ng-monitor) in tiup-playground (#1648, @nexustar)
• Add support of inserting custom scrape_configs to Prometheus configs in tiup-cluster (#1641, @nexustar)
• [experimental] Support 2-staged scaling out for tiup-cluster (#1638 #1642, @srstack)
  • Scaling out of a TiDB cluster can be devided with --stage1 and --stage2 arguments, the stage 1 deploys files and configs but not starting the new instances, and the stage 2 actually starts the new instances and reload necessary configs
  • This could be useful if you want to modify config of the new instances or use a custom binary with patch before the first start of the new instances
• [experimental] Implement plain text output and support custom output writer for logs (#1646, @AstroProfundis)

Fixes

• Fix incorrect progress bar displaying in some tasks (#1624, @nexustar)
• Fix incorrect argument flags in tiup-playground (#1635, @srstack)
• Fix files of monitoring agents and TiDB audit log not cleaned with clean command of tiup-cluster (#1643 #1644, @srstack)
• Fix confirmation prompt in scale-out can not be skipped with --yes argument in tiup-cluster (#1645, @srstack)
• Fix directory conflict error in some circumstances even when node is marked as ignore_exporter (#1649, @AstroProfundis)
• Fix DM nodes not added to node_exporter target list in Prometheus config in tiup-dm (#1654, @AstroProfundis)

Improvements

• Add significant warning when --force argument is set for scale-in command in tiup-cluster (#1629, @AstroProfundis)
• Add environment variables to skip topology sanity check in scale-in command in tiup-cluster (#1627, @AstroProfundis)
• Update examples to use --without-monitor instead of --monitor for tiup-playground (#1639, @dveeden)