gpu_sysman: Give variant config examples in doc

With the given 2 example variant configs, one does not
miss any of the information that Sysman might provide.

Signed-off-by: Eero Tamminen <eero.t.tamminen@intel.com>

diff --git a/src/collectd.conf.pod b/src/collectd.conf.pod
index d9f1236a0d1e6773462f8c61fc93a07a5ff606dd..00c270fa607f56e010bd5320ac3b40dc6ca922a4 100644 (file)
--- a/src/collectd.conf.pod
+++ b/src/collectd.conf.pod
@@ -3736,10 +3736,10 @@ Several of the metric types support multiple variants for their
 values. This option specifies which ones of them are to be reported.
 
 Default is to report all variants ("counter:rate:ratio"). To reduce
-amount of data, it is better to configure just the relevant one for
-given use. Note that some of the metric types support only two of
-these variants, whereas metrics supporting only single variant ignore
-this option.
+amount of data, it is better to configure just the relevant ones for
+given use (e.g. "counter:ratio" or "rate:ratio"). Note that some of
+the metric types support only two of these variants, whereas metrics
+supporting only single variant ignore this option.
 
 Counter metric variant (e.g. HW energy usage as Joules counter) is
 preferred by Prometheus as doing rate calculations in Prometheus is