CHANGELOG.md

3.0.1 / 2024-03-21

• add fileset label to user and group quotas to avoid duplicates (#68)

3.0.0 / 2023-11-20

• [BREAKING] Change how mmhealth event filtering works and make events a metric (#62)
  • Add gpfs_health_event metric.
  • The collector.mmhealth.ignored-event flag will only filter events for the gpfs_health_event metric
• TLS and auth support (#65)

3.0.0-rc.1 / 2023-07-12

• [BREAKING] Change how mmhealth event filtering works and make events a metric (#62)
  • Add gpfs_health_event metric.
  • The collector.mmhealth.ignored-event flag will only filter events for the gpfs_health_event metric

2.6.0 / 2023-07-09

• Add users and groups mmrepquota (#59)

2.5.1 / 2023-07-05

• Skip NODE mmhealth status if filtering out releated event (#61)

2.5.0 / 2023-06-20

• Support filtering mmhealth by events (#60)

2.4.0 / 2023-05-06

• Update to Go 1.20.3 and update Go dependencies (#58)

2.3.0 / 2023-04-30

• Add mmlsqos collector (#56)

2.2.0 / 2022-10-13

• Allow sudo command to be changed (#54)

2.1.0 / 2022-09-19

• Update mmdf collector to collect pool data

2.0.0 / 2022-03-08

• [BREAKING] Change how waiter metrics are presented and stored
  • Replace gpfs_mmdiag_waiter with gpfs_waiter_seconds that is a histogram, no longer use thread label
  • Replace gpfs_mmdiag_waiter_info with gpfs_waiter_info_count that is count of waiter name without thread or reason labels
  • Add flag --collector.waiter.buckets to define histogram buckets
  • Add flag --collector.waiter.log-reason to enable logging of waiter reasons
• [BREAKING] Rename mmdiag collector to waiter
  • Replace --collector.mmdiag with --collector.waiter
  • Replace --no-collector.mmdiag with --no-collector.waiter
  • Remove --collector.mmdiag.waiter-threshold flag
  • Replace --collector.mmdiag.waiter-exclude with --collector.waiter.exclude
  • Replace --collector.mmdiag.timeout with --collector.waiter.timeout
• [BREAKING] The waiter exclude will only compare against waiter name
• Update Go to 1.17
• Update Go module dependencies

2.0.0-rc.2 / 2021-09-23

• [BREAKING] Change how waiter metrics are presented and stored
  • Replace gpfs_mmdiag_waiter with gpfs_waiter_seconds that is a histogram, no longer use thread label
  • Replace gpfs_mmdiag_waiter_info with gpfs_waiter_info_count that is count of waiter name without thread or reason labels
  • Add flag --collector.waiter.buckets to define histogram buckets
  • Add flag --collector.waiter.log-reason to enable logging of waiter reasons
• [BREAKING] Rename mmdiag collector to waiter
  • Replace --collector.mmdiag with --collector.waiter
  • Replace --no-collector.mmdiag with --no-collector.waiter
  • Remove --collector.mmdiag.waiter-threshold flag
  • Replace --collector.mmdiag.waiter-exclude with --collector.waiter.exclude
  • Replace --collector.mmdiag.timeout with --collector.waiter.timeout
• [BREAKING] The waiter exclude will only compare against waiter name

1.5.1 / 2021-06-17

• Fix mmdf collector to still write last collection metric during errors

1.5.0 / 2021-06-07

• Add gpfs_mmdiag_waiter_info metric

1.4.0 / 2021-05-27

• Add config collector, enabled by default

1.3.0 / 2021-04-23

Changes

• Update to Go 1.16
• Add mmlsfileset collector

1.2.0 / 2021-04-15

Changes

• Add mmlssnapshot collector

1.1.2 / 2021-04-12

Bug fixes

• Do not produce errors if no metadata is reported by mmdf

1.1.1 / 2021-03-31

Bug fixes

• Fix possible index out of range parsing errors with mmdf collector

1.1.0 / 2021-01-02

Changes

• Allow mmhealth items to be filtered out via CLI flags
• Allow mmces services to be filtered out via CLI flags

1.0.0 / 2020-11-24

Breaking Changes

• Remove --exporter.use-cache flag and all caching logic
• Rename several metrics to standardize naming conventions
  • gpfs_perf_read_bytes to gpfs_perf_read_bytes_total
  • gpfs_perf_write_bytes to gpfs_perf_write_bytes_total
  • gpfs_perf_operations to gpfs_perf_operations_total
  • gpfs_fs_inodes_allocated to gpfs_fs_allocated_inodes
  • gpfs_fs_inodes_free to gpfs_fs_free_inodes
  • gpfs_fs_inodes_total to gpfs_fs_total_inodes
  • gpfs_fs_inodes_used to gpfs_fs_used_inodes
  • gpfs_fs_total_inodes to gpfs_fs_inodes
  • gpfs_fs_total_bytes to gpfs_fs_size_bytes
  • gpfs_fs_metadata_total_bytes to gpfs_fs_metadata_size_bytes
• Removed metrics that can be calculated using other metrics
  • gpfs_fs_metadata_free_percent
  • gpfs_fs_free_percent
• Remove nodename label from gpfs_perf_* metrics, replace with gpfs_perf_info metric
• mmces state metrics will have one metric per possible state, with active state having value 1
• mmhealth status metrics will have one metric per possible status with active status having value 1

Changes

• Update to Go 1.15 and update all dependencies
• Improved error handling for cron gpfs_mmdf_exporter
• Add mmrepquota collector to collect quota information for filesets

0.11.1 / 2020-04-21

• Fix mount collector to avoid false positives

0.11.0 / 2020-04-04

• Improve timeout/error handling around mmlsfs and add tests

0.10.0 / 2020-04-04

• Simplified timeout and error handling

0.9.0 / 2020-03-16

Changes

• Allow caching of metrics if errors or timeouts occur
• Improved testing

0.8.0 / 2020-03-05

Changes

• Add mmgetstate collector and metrics
• Use promlog for logging

0.7.0 / 2020-03-02

Changes

• Add timeouts to all collectors

0.6.0 / 2020-02-25

Changes

• Update client_golang dependency
• Testing improvements

0.5.0 / 2020-02-18

Changes

• Support excluding waiters

0.4.0 / 2020-02-17

Changes

• Refactor mmdiag waiter metric collection

0.3.1 / 2020-02-17

Fixes

• Avoid duplicate metrics for collector errors

0.3.0 / 2020-02-15

Changes

• Add mmdiag collector with waiters metric
• Add mmces collector with service state metrics

0.2.0 / 2020-01-30

Changes

• Move all metrics to /metrics endpoint, remove /gpfs endpoint
• Add --web.disable-exporter-metrics flag

0.1.0 / 2020-01-29

Changes

• Rename gpfs_mmhealth_state to gpfs_mmhealth_status
• Add status label to mmhealth status metric

0.0.1 / 2020-01-29

Changes

• Initial Release