ports/iperf3
ports/iperf3
1
1
Fork 0
Code Releases Activity

Towards issue #278. Don't print small, dataless last intervals. (#592)

Browse Source 
This is mostly a cosmetic fix...due to timing differences or
interference from data packets, it is possible that the server
will have a very small last measurement window, possibly with
no data transferred.  This looks odd (although it's not incorrect)
and we'd like to eliminate these.  If there's an interval
(presumably it'd be the last measurement interval) less than
10% of the length of a standard measurement/reporting interval,
and there are no bytes transferred in the interval, then suppress
printing the results and don't add them to the grand total.
This commit is contained in:
Bruce A. Mah 2017-06-05 10:38:36 -07:00 committed by GitHub
parent 7f06146221
commit b5597278b5
1 changed files with 44 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

44
src/iperf_api.c
View File

@ -2459,6 +2459,50 @@ iperf_print_intermediate(struct iperf_test *test)
int total_packets = 0, lost_packets = 0;
double avg_jitter = 0.0, lost_percent;
/*
* Due to timing oddities, there can be cases, especially on the
* server side, where at the end of a test there is a fairly short
* interval with no data transferred. This could caused by
* the control and data flows sharing the same path in the network,
* and having the control messages for stopping the test being
* queued behind the data packets.
*
* We'd like to try to omit that last interval when it happens, to
* avoid cluttering data and output with useless stuff.
* So we're going to try to ignore very short intervals (less than
* 10% of the interval time) that have no data.
*/
int interval_ok = 0;
SLIST_FOREACH(sp, &test->streams, streams) {
irp = TAILQ_LAST(&sp->result->interval_results, irlisthead);
if (irp) {
double interval_len = timeval_diff(&irp->interval_start_time,
&irp->interval_end_time);
if (test->debug) {
printf("interval_len %f bytes_transferred %lu\n", interval_len, irp->bytes_transferred);
}
/*
* If the interval is at least 10% the normal interval
* length, or if there were actual bytes transferrred,
* then we want to keep this interval.
*/
if (interval_len >= test->stats_interval * 0.10 ||
irp->bytes_transferred > 0) {
interval_ok = 1;
if (test->debug) {
printf("interval forces keep\n");
}
}
}
}
if (!interval_ok) {
if (test->debug) {
printf("ignoring short interval with no data\n");
}
return;
}
if (test->json_output) {
json_interval = cJSON_CreateObject();
if (json_interval == NULL)