github/netdata
7
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Update README.md

Browse Source
This commit is contained in:
Costa Tsaousis 2023-10-27 06:19:39 +03:00 committed by GitHub
parent bd8dc71c29
commit 4071d1a3ff
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
1 changed files with 2 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
README.md
View File

@ -421,6 +421,8 @@ Yes, you can!
Netdata has been designed to spread disk writes across time. Each metric is flushed to disk every 17 minutes, but metrics are flushed evenly across time, at an almost constant rate. Also, metrics are packed into bigger blocks we call `extents` and are compressed with LZ4 before saving them, to minimize the number of I/O operations made.
Netdata also employs direct I/O for all its database operations, ensuring optimized performance. By managing its own caches, Netdata avoids overburdening system caches, facilitating a harmonious coexistence with other applications.
Single node Agents (not Parents), should have a constant rate of about 50 KiB/s or less, with some spikes above that every minute (flushing of tier 1) and higher spikes every hour (flushing of tier 2).
Health Alerts and Machine-Learning run queries to evaluate their expressions and learn from the metrics' patterns. These are also spread over time, so there should be an almost constant read rate too.