archive/netdata_netdata
0
0
Fork 0
mirror of https://github.com/netdata/netdata.git synced 2025-05-07 02:30:12 +00:00
Code Issues Projects Releases Wiki Activity
netdata_netdata/docs/observability-centralization-points
History
Costa Tsaousis a399128dbf
config parsers () 
* added parser for durations

* preliminary work for timeframes

* Update CMakeLists.txt

* updated parsing and generation for durations

* renames

* report parser errors; added compatibility to existing config_parse_duration()

* duration parsing is used on most netdata.conf and stream.conf entries

* more uses of duration parsing; simplification of stream.conf

* code cleanup

* more duration changes

* added html playground

* improved js code

* duration parsing applied to dbengine retention

* fixed doc

* simplified logic; added size parser

* added parsing for sizes

* renames and documentation updates

* hide appconfig internals from the rest of netdata

* fix crash on cleanup of streaming receivers

* fix buffer overflow in gorilla compression

* config return values are const

* ksm set to auto

* support reformatting migrated values

* removed obsolete metrics correlations settings

* split appconfig to multiple files

* durations documentation

* sizes documentation

* added backward compatibility in retention configuration

* provide description on migrations and reformattings

* config options are now a double linked list

* config sections are now a double linked list; config uses spinlocks; code cleanup and renames

* added data type to all config options

* update data types

* split appconfig api to multiple files

* code cleanup and renames

* removed size units above PiB

* Revert "fix buffer overflow in gorilla compression"

This reverts commit 3d5c48e84b.

* appconfig internal api changes
2024-09-04 14:42:01 +03:00
..
logs-centralization-points-with-systemd-journald Update README.md 2024-03-17 23:21:02 +02:00
metrics-centralization-points config parsers () 2024-09-04 14:42:01 +03:00
best-practices.md Documentation changes, new files and restructuring the hierarchy () 2024-02-20 09:08:46 +02:00
README.md Documentation changes, new files and restructuring the hierarchy () 2024-02-20 09:08:46 +02:00

README.md

Observability Centralization Points

Netdata supports the creation of multiple independent Observability Centralization Points, aggregating metric samples, logs and metadata within an infrastructure.

Observability Centralization Points are crucial for ensuring comprehensive monitoring and observability across an infrastructure, particularly under the following conditions:

  1. Ephemeral Systems: For systems like Kubernetes nodes or ephemeral VMs that may not be persistently available, centralization points ensure that metrics and logs are not lost when these systems go offline. This is essential for maintaining historical data for analysis and troubleshooting.

  2. Resource Constraints: In scenarios where the monitored systems lack sufficient resources (disk space or I/O bandwidth, CPU, RAM) to handle observability tasks effectively, centralization points offload these responsibilities, ensuring that production systems can operate efficiently without compromise.

  3. Multi-node Dashboards without Netdata Cloud: For environments requiring aggregated views across multiple nodes but without the use of Netdata Cloud, Netdata Parents can aggregate this data to provide comprehensive dashboards, similar to what Netdata Cloud offers.

  4. Netdata Cloud Access Restrictions: In cases where monitored systems cannot connect to Netdata Cloud (due to a firewall policy), a Netdata Parent can serve as a bridge, aggregating data and interfacing with Netdata Cloud on behalf of these restricted systems.

When multiple independent centralization points are available:

  • Netdata Cloud queries all of them in parallel, to provide a unified infrastructure view.

  • Without Netdata Cloud, the dashboards of each of the Netdata Parents provide unified views of the infrastructure aggregated to each of them (metrics and logs).