MyGit

v1.46.0

netdata/netdata

版本发布时间: 2024-06-19 23:45:59

netdata/netdata最新发布版本:v1.47.2(2024-09-24 21:54:58)

Table of Contents

Netdata Growth

Netdata is being downloaded more than 1.5 million times per day, as reported by Cloudflare, which distributes our binary packages, and Docker Hub, which distributes our docker containers!

Thanks to your love ❤️, Netdata is leading the observability category in CNCF, having significantly more stars than Elasticsearch, Grafana, Prometheus and all other observability solutions listed in the CNCF landscape.

We are committed to providing the most advanced and innovative observability solution, to help us minimize monitoring costs while providing AI-powered high-fidelity monitoring!

Do you like Netdata? Give Netdata a ⭐ too, on GitHub!

Release Summary

Another 3 months have passed, we're already mid-way through 2024, and it's time for another Netdata release! The Netdata team has been super busy cooking up some great new features we're sure you will enjoy. Users can now dynamically configure collectors and alerts from the UI, the long-awaited Native Windows agent is almost here, AWS billing integration and so many more cool features.

In a crowded observability market, Netdata stands out with real-time monitoring and a truly unique decentralized architecture which ensures the data, always stays on-prem!

Release Highlights

Dynamic Configuration out of beta

Manage Netdata from the Dashboard:

After fine-tuning of our Dynamic Configuration feature (DynCfg) based on extensive feedback, we have now made a GA release available to our Business and Homelab plans.

Netdata subscriptions on AWS Marketplace

Subscribe to Netdata from the AWS Marketplace AWS Customers can now manage their payments and subscriptions directly from AWS - no need for extra budget approvals or billing complexities!

image

Native Windows Support

image

Netdata's current code base can now be compiled on Windows and runs smoothly on all versions of the Windows operating system.

Currently, this feature is in the internal testing phase. We are working on packaging it and adding Windows-specific collectors. In the next Netdata release, we will provide binary packages for Microsoft Windows.

Okta and OpenID SSO Authentication

image

Effortlessly manage your teams’ access with SSO through Okta or OpenID integrations. This feature ensures streamlined access and heightened security across your systems. We are now looking into building the next step to allow you to manage your users' Roles and Permissions!

Time Based Retention

Netdata now offers a more granular approach to data retention, allowing you to manage storage based on both time and disk space. This update provides greater control and helps you optimize storage usage for your specific needs.

Tier Resolution Time Limit Size Limit
0 high 14 days 1 GiB
1 middle 3 months 1 GiB
2 low 2 years 1 GiB

With these defaults, Netdata requires approximately 4 GiB of storage space (including metadata).

Retention Settings

You can fine-tune retention for each tier by setting either the time limit or size limit to zero. This enables various retention strategies as shown in the table below:

Setting Retention Behavior
Size Limit = 0, Time Limit > 0 Time-based only: data is stored for a specific duration regardless of disk usage.
Time Limit = 0, Size Limit > 0 Space-based only: data is stored until it reaches a certain amount of disk space, regardless of time.
Time Limit > 0, Size Limit > 0 Combined time and space limits: data is deleted once it reaches either the time limit or the disk space limit, whichever comes first.

Monitoring Retention Utilization

A new chart has been added to the Netdata Monitoring section of your dashboard. This chart displays the utilization percentage for both time and space limits across all tiers, providing a clear overview of your storage usage.

Database Compression

Netdata now utilizes ZSTD compression within its database engine (dbengine), significantly reducing the storage space required for your monitoring data. This translates to a mere 0.5 bytes per sample on disk for the high-resolution tier!

Netdata's impressive compression capabilities are the result of a powerful combination of features:

ClickHouse Collector

image

Netdata now ships with a native ClickHouse collector, which collects performance data from ClickHouse for connections, queries, resources, replication, IO, and data operations (inserts, selects, merges) using HTTP requests and ClickHouse system tables.

OpenTelemetry Compatibility

All the units on our charts and metrics are now compatible with the Unified Code for Units of Measure (UCUM). This alignment ensures precision, consistency, and ease of integration across diverse systems, paving the way for more seamless observability and monitoring experiences. Stay tuned for more updates as we continue to enhance our platform!

More Notification Integrations

image

And as always we keep adding more integrations for notifications. You can now send your Netdata alerts from Netdata Cloud to Splunk, VictorOps, MS Teams, and more!

Acknowledgments

Contributions

Collectors

New/Rewritten
  • Add collector for ClickHouse database (go.d/clickhouse) (#17743, @ilyam8)
  • Add collector for DMCache (#17947, @ilyam8)
  • Add collector for Rspamd spam filtering system (go.d/rspamd) (#17679, @ilyam8)
  • Add collector for StoreCLI RAIDs (go.d/storcli) (#17454, @ilyam8)
  • Add collector for LVM thin volumes (go.d/lvm) (#17394, @ilyam8)
  • Add collector for ZFS pools (go.d/zfspool) (#17367, @ilyam8)
  • Add collector for Intel GPUs (go.d/intelgpu) (#17344, @ilyam8)
  • Rewrite collector for LiteSpeed in golang (go.d/litespeed) (#17665, @ilyam8)
  • Rewrite collector for HPE Smart Arrays in golang (go.d/hpssa) (#17637, @ilyam8)
  • Rewrite collector for SMART metrics in golang (go.d/smartctl) (#17536, @ilyam8)
  • Rewrite collector for Fail2Ban in golang (go.d/fail2ban) (#17501, @ilyam8)
  • Rewrite collector for hardware sensors in golang (go.d/sensors) (#17466, @ilyam8)
  • Rewrite collector for HDD temperatures in golang (go.d/hddtemp) (#17462, @ilyam8)
  • Rewrite collector for Adapter RAIDs in golang (go.d/adaptecraid) (#17428, @ilyam8)
  • Rewrite collector for MegaCLI RAIDs in golang (go.d/megacli) (#17410, @ilyam8)
Improvements
  • Add initial support for mpt3sas controllers (go.d/storcli) (#17938, @ilyam8)
  • Try requesting extended data if no expiration date (go.d/whoisquery) (#17913, @ilyam8)
  • Do not use "/dev" to find Zram devices (proc/sys_block_zram) (#17900, @ilyam8)
  • Add an option to select specific GPU for monitoring (go.d/intelgpu) (#17884, @ilyam8)
  • Remove unused options for a cleaner configuration (tcp.plugin, proc/proc_net_dev, cgroups.plugin) (#17835, #17837, #17795, @ilyam8)
  • Add alerts for LVM thins pools (go.d/lvm) (#17807, @ilyam8)
  • Add alerts for ClickHouse (go.d/clickhouse) (#17760, @ilyam8)
  • Add more metrics to ClickHouse collector (go.d/clickhouse) (#17756, @ilyam8)
  • Add running queries metric to ClickHouse collector (go.d/clickhouse) (#17751, @ilyam8)
  • Add an option to skip transient systemd units (go.d/systemdunits) (#17725, @ilyam8)
  • Ignore HTTP 101 (Switching Protocols) requests when calculating request processing times (go.d/weblog) (#17717, @ilyam8)
  • Add dyncfg action to render configurations in YAML format (go.d.plugin dyncfg) (#17684, @ilyam8)
  • Add Rspamd to apps_groups.conf (#17674, @ilyam8)
  • Improve locating of user and stock configuration files (go.d.plugin) (#17661, @ilyam8)
  • Add Valkey to apps_groups.conf (#17639, @mohd-akram)
  • Add alerts for ISC DHCP pools (go.d/isc_dhcp) (#17634, @ilyam8)
  • Reorganize charts: create a separate chart for each DHCP pool instead of grouping them as dimensions within a single chart (go.d/isc_dhcp) (#17629, @ilyam8)
  • Reorganize charts: create a separate chart for each file/directory instead of grouping them as dimensions within a single chart (#17624, @ilyam8)
  • Adjust the return code from 405 to 200 when enabling running jobs (go.d.plugin dyncfg) (#17611, @ilyam8)
  • Add systemd unit files state metric (go.d/systemdunits) (#17606, @ilyam8)
  • Add an option to filter devices and enhance unit detection logic (go.d/smartctl) (#17565, @ilyam8)
  • Add heuristics to automatically determine the units of SMART attributes (go.d/smartctl) (#17564, @ilyam8)
  • Improve unit detection logic for snmp_exporter (go.d/prometheus) (#17524, @ilyam8)
  • Use XML format instead of CSV by default (go.d/nvidia_smi) (#17481, @ilyam8)
  • Add alerts for Storcli RAIDs and add BBU temperature charts (go.d/storcli) (#17460, @ilyam8)
  • Switch from using nvme binary to ndsudo (go.d/nvme) (#17386, @ilyam8)
  • Switch from using intel_gpu_top binary to ndsudo (go.d/intelgpu) (#17380, @ilyam8)
  • Enable Intel GPU collector by default and update its metadata (go.d/intelgpu) (#17370, @ilyam8)
  • Rename label engine to engine_class (go.d/intelgpu) (#17361, @ilyam8)
  • Add intel_gpu_top to apps_groups.conf (#17354, @ilyam8)
  • Add power consumption metric to Linux Power Supply collector (proc/sys_class_power_supply) (#17329, @eyusupov)
  • Adjust SWAP module for changes in kernels 6.8+ (ebpf.plugin) (#17295, @thiagoftsm)
  • Stop using kf_sock_inpcb on FreeBSD 14+ (apps.plugin) (#17243, @glebius)
  • Add instance label to cgroup and application group charts (ebpf.plugin) (#17156, @thiagoftsm)
Bug fixes
  • Fix missing config schema (go.d/phpfpm) (#17872, @ilyam8)
  • Fix systemd unit files selector (go.d/systemdunits) (#17622, @ilyam8)
  • Fix the process handling to properly wait for intel_gpu_top to exit (go.d/intelgpu) (#17384, @ilyam8)
  • Fix wrong context of the Megacli PD media errors alert (go.d/megacli) (#17419, @ilyam8)
  • Fix the issue where not all Netdata log levels were respected (go.d.plugin, python.d.plugin) (#17587, @ilyam8)
  • Fix the absence of connection read/write timeouts (go.d/phpfpm) (#17784, @ilyam8)
  • Fix table bloat stats calculation (go.d/postgres) (#17598, @ilyam8)
  • Fix the issue where single quotes in labels caused charts to break (go.d/prometheus) (#17570, @ilyam8)
  • Fix the plugin's attempt to update a dimension without prior creation (go.d/traefik) (#17511, @ilyam8)
  • Fix parsing of Physical Drives information (python.d/hpssa) (#17294, @Metric-Void)
  • Fix buffer overflow in the function that detects cgroups version (cgroups.plugin) (#17799, @ilyam8)
  • Fix creating of Pressure charts when metrics are not available (cgroups.plugin) (#17800, @ilyam8)
  • Fix ClickHouse detection in apps_groups.conf (#17742, @ilyam8)
  • Fix swap context of cgroup charts (ebpf.plugin) (#17740, @ilyam8)
  • Fix calculation of last read and write sectors (proc/diskstats) (#17718, @ilyam8)
  • Fix indentation in the MongoDB collector configuration (go.d/mongodb) (#17715, @louis-lau)
  • Fix the issue with identifying Rspamd in network listeners service discovery (go.d.plugin) (#17686, @ilyam8)
  • Fix missing settings in the Ping collector configuration schema (go.d/ping) (#17642, @ilyam8)
  • Fix issue preventing metrics collection for Intel GPU when update_every setting is greater than 3 seconds (go.d/intelgpu) (#17491, @ilyam8)
  • Fix 'family' value of Linux Power Supply charts (proc/sys_class_power_supply) (#17338, @ilyam8)
  • Fix missing "body" and "method" settings in modules configuration schema (go.d.plugin) (#17325, @ilyam8)
  • Fix context of the Cache battery status chart (go.d/hpssa) (#17654, @ilyam8)
  • Fix unit of some ebpf charts (ebpf.plugin) (#17809, @ilyam8)
  • Fix duplicate chart context (#17267, @ilyam8)
  • Fix the issue where cgroup OOMKills charts were not removed when the cgroup was deleted (ebpf.plugin) (#17734, @thiagoftsm)
Other

Packaging/Installation

All changes

Documentation

All changes

Other Notable Changes

New Features
  • Add a DYNCFG action to transform user-created configurations into on-disk formats compatible with the chosen plugin (#17692, @ktsaou)
  • Add netdatacli command to remove stale nodes (#17691, @stelfrag)
  • Add support for time-based retention to dbengine (#17413, @stelfrag)
  • Add support for ZSTD compression to dbengine (#17244, @ktsaou)
Improvements
Other

Deprecation notice

Changed in this release: Database disk space usage

Default disk space limits have been increased to better accommodate the growing data collection needs of typical Netdata deployments. In this release, the default disk space limit is increased to 1 Gigabyte (GB) for each tier, resulting in a total of 3 GB.

Old vs. New Disk Space Limits

Tier Old Limit New Limit
0 256 MB 1 GB
1 128 MB 1 GB
2 64 MB 1 GB

You can manually adjust the disk space limits to your preferences in the netdata.conf configuration file.

Netdata now offers time-based retention alongside the existing disk-based retention.

Changed in this release: Rewritten/Changed collectors

Rewritten Collectors: we have rewritten a number of collectors from Python (python.d.plugin) to Go (go.d.plugin).

Deprecated Collector (Python) Replacement Collector (Go)
python.d/litespeed go.d/litespeed
python.d/hpssa go.d/hpssa
python.d/smartd_log go.d/smartctl
python.d/fail2ban go.d/fail2ban
python.d/sensors go.d/sensors
python.d/hddtemp go.d/hddtemp
python.d/adaptec_raid go.d/adaptec_raid
python.d/megacli go.d/megacli

Reorganized Charts: charts for specific collectors have been reorganized to provide a more intuitive user experience:

Collector Change
go.d/isc_dhcp charts for DHCP metrics are now separated by individual DHCP pools instead of being grouped as dimensions within a single chart
go.d/filecheck charts for various file/directory metrics are now separated by individual files or directories, rather than being grouped as dimensions within a single chart

Important Note: While most users won't experience any disruption, there's one key point to consider if you're using Netdata solely as a metric collector and exporting data to an external Time-Series Database (TSDB). Due to the rewrite, some metric names have been changed. If you were previously relying on these specific metric names to create charts or alerts within your TSDB, you may need to update your configurations to reflect the new names.

To view a complete list of all metrics collected by a specific collector, refer to its documentation.

Netdata Release Meetup

Join the Netdata team on the 27th of June at 16:30 UTC for the Netdata Release Meetup.

Together we’ll cover:

RSVP now - we look forward to meeting you.

Support options

As we grow, we stay committed to providing the best support ever seen from an open-source solution. Should you encounter an issue with any of the changes made in this release or any feature in the Netdata Agent, feel free to contact us through one of the following channels:

相关地址:原始地址 下载(tar) 下载(zip)

1、 integrations.js 4.61MB

2、 netdata-aarch64-latest.gz.run 54.05MB

3、 netdata-aarch64-v1.46.0.gz.run 54.05MB

4、 netdata-armv6l-latest.gz.run 52.92MB

5、 netdata-armv6l-v1.46.0.gz.run 52.92MB

6、 netdata-armv7l-latest.gz.run 51.87MB

7、 netdata-armv7l-v1.46.0.gz.run 51.87MB

8、 netdata-latest.gz.run 65.82MB

9、 netdata-latest.tar.gz 71.65MB

10、 netdata-ppc64le-latest.gz.run 55.28MB

11、 netdata-ppc64le-v1.46.0.gz.run 55.28MB

12、 netdata-v1.46.0.gz.run 65.82MB

13、 netdata-v1.46.0.tar.gz 71.65MB

14、 netdata-x86_64-latest.gz.run 65.82MB

15、 netdata-x86_64-v1.46.0.gz.run 65.82MB

16、 sha256sums.txt 1.48KB

查看:2024-06-19发行的版本