Update info on plugins (#5101)
Also add a bit more info on memory usage, due to #5098
Chris Akritidis committed
Jan 23, 2019 at 15:54 UTC
2c01d9f5165e6c0a37414bf53898f9027dd2b5c4
1 file changed
+6
-17
docs/Performance.md
+6
-17
@@ -110,6 +110,7 @@ Edit `/etc/netdata/netdata.conf`, find the `[plugins]` section:
110
apps = no
111
charts.d = no
112
node.d = no
113
+ python.d = no
114
115
plugins directory = /usr/libexec/netdata/plugins.d
116
enable running new plugins = no
@@ -128,6 +129,7 @@ plugin|description
129
`apps`|a plugin that monitors system processes. It is very complex and heavy (consumes twice the CPU resources of the netdata daemon), so if you don't need to monitor the process tree, you can disable it.
130
`charts.d`|BASH plugins (squid, nginx, mysql, etc). This is a heavy plugin, that consumes twice the CPU resources of the netdata daemon.
131
`node.d`|node.js plugin, currently used for SNMP data collection and monitoring named (the name server).
132
+`python.d`|has many modules and can use over 20MB of memory.
133
134
For most IoT devices, you can disable all plugins except `proc`. For `proc` there is another section that controls which functions of it you need. Check the next section.
135
@@ -135,30 +137,17 @@ For most IoT devices, you can disable all plugins except `proc`. For `proc` ther
137
138
### 2. Disable internal plugins
139
138
-In this section you can select which modules of the `proc` plugin you need. All these are run in a single thread, one after another. Still, each one needs some RAM and consumes some CPU cycles.
140
+In this section you can select which modules of the `proc` plugin you need. All these are run in a single thread, one after another. Still, each one needs some RAM and consumes some CPU cycles. With all the modules enabled, the `proc` plugin adds ~9 MiB on top of the 5 MiB required by the netdata daemon.
141
142
```
143
[plugin:proc]
144
# /proc/net/dev = yes # network interfaces
145
# /proc/diskstats = yes # disks
144
- # /proc/net/snmp = yes # generic IPv4
145
- # /proc/net/snmp6 = yes # generic IPv6
146
- # /proc/net/netstat = yes # TCP and UDP
147
- # /proc/net/stat/conntrack = yes # firewall
148
- # /proc/net/ip_vs/stats = yes # IP load balancer
149
- # /proc/net/stat/synproxy = yes # Anti-DDoS
150
- # /proc/stat = yes # CPU, context switches
151
- # /proc/meminfo = yes # Memory
152
- # /proc/vmstat = yes # Memory operations
153
- # /proc/net/rpc/nfsd = yes # NFS Server
154
- # /proc/sys/kernel/random/entropy_avail = yes # Cryptography
155
- # /proc/interrupts = yes # Interrupts
156
- # /proc/softirqs = yes # SoftIRQs
157
- # /proc/loadavg = yes # Load Average
158
- # /sys/kernel/mm/ksm = yes # Memory deduper
159
- # netdata server resources = yes # netdata charts
146
+...
147
```
148
149
+Refer to the [proc.plugins documentation](../collectors/proc.plugin/) for the list and description of all the proc plugin modules.
150
+
151
### 3. Lower internal plugin update frequency
152
153
If netdata is still using a lot of CPU, lower its update frequency. Going from per second updates, to once every 2 seconds updates, will cut the CPU resources of all netdata programs **in half**, and you will still have very frequent updates.