summaryrefslogtreecommitdiffstats
path: root/collectors/REFERENCE.md
diff options
context:
space:
mode:
authorvkalintiris <vasilis@netdata.cloud>2022-11-22 04:52:15 +0200
committerGitHub <noreply@github.com>2022-11-22 04:52:15 +0200
commit2d5f3acf71f0c759056a3269987fee484566bc4c (patch)
tree5246e1080ea721ba84e5f749f8d8e98d978d81c8 /collectors/REFERENCE.md
parent147552807bc19af949fe3cb315c4743dadfa7f0b (diff)
Do not force internal collectors to call rrdset_next. (#13926)
* Remove calls to rrdset_next(). * Rm checks plugin * Update documentantion * Call rrdset_next from within rrdset_done This wraps up the removal of rrdset_next from internal collectors, which removes a lot of unecessary code and the need for if/else clauses in every place. The pluginsd parser is the only component that calls rrdset_next*() functions because it's not strictly speaking a collector but more of a collector manager/proxy. With the current changes it's possible to simplify the API we expose from RRD significantly, but this will be follow-up work in the future. * Remove stale reference to checks.plugin * Fix RRD unit test rrdset_next is not meant to be called from these tests. * Fix db engine unit test. * Schedule rrdset_next when we have completed at least one collection. * Mark chart creation clauses as unlikely. * Add missing brace to fix FreeBSD plugin.
Diffstat (limited to 'collectors/REFERENCE.md')
-rw-r--r--collectors/REFERENCE.md5
1 files changed, 0 insertions, 5 deletions
diff --git a/collectors/REFERENCE.md b/collectors/REFERENCE.md
index 949858f600..939b189ee7 100644
--- a/collectors/REFERENCE.md
+++ b/collectors/REFERENCE.md
@@ -148,11 +148,6 @@ collect_data() {
// attach a metric to it
rd = rrddim_add(st, "id", "name", multiplier, divider, algorithm);
}
- else {
- // this chart is already created
- // let Netdata know we start a new iteration on it
- rrdset_next(st);
- }
// give the collected value(s) to the chart
rrddim_set_by_pointer(st, rd, collected_value);