neels has uploaded this change for review. ( https://gerrit.osmocom.org/c/osmo-hnbgw/+/37188?usp=email )

Change subject: nft-kpi: remove X34 drifting: adjust delay by elapsed time ......................................................................

nft-kpi: remove X34 drifting: adjust delay by elapsed time

Change-Id: I04f572890c04a48bb19c59f613a492ef96624baa --- M include/osmocom/hnbgw/hnbgw.h M src/osmo-hnbgw/nft_kpi.c 2 files changed, 48 insertions(+), 6 deletions(-)

git pull ssh://gerrit.osmocom.org:29418/osmo-hnbgw refs/changes/88/37188/1

diff --git a/include/osmocom/hnbgw/hnbgw.h b/include/osmocom/hnbgw/hnbgw.h index 36358b1..a647abd 100644 --- a/include/osmocom/hnbgw/hnbgw.h +++ b/include/osmocom/hnbgw/hnbgw.h @@ -499,6 +499,7 @@ struct { bool active; struct osmo_timer_list get_counters_timer; + struct timespec next_timer; } nft_kpi; };

diff --git a/src/osmo-hnbgw/nft_kpi.c b/src/osmo-hnbgw/nft_kpi.c index 4fa932f..aff981f 100644 --- a/src/osmo-hnbgw/nft_kpi.c +++ b/src/osmo-hnbgw/nft_kpi.c @@ -690,15 +690,47 @@ /* main thread */ static void nft_kpi_get_counters_schedule(void) { - unsigned long period_s; + struct timespec now; + struct timespec period; + struct timespec diff; + struct timespec *next = &g_hnbgw->nft_kpi.next_timer; unsigned long period_us = osmo_tdef_get(hnbgw_T_defs, -34, OSMO_TDEF_US, 1000000); - if (period_us < 1) - period_us = 1; - period_s = period_us / 1000000; - period_us %= 1000000;

+ period.tv_sec = period_us / 1000000; + period.tv_nsec = (period_us % 1000000) * 1000; + + /* Try to keep the period of getting counters close to the configured period, i.e. don't drift by the time it + * takes to read counters. */ + osmo_clock_gettime(CLOCK_MONOTONIC, &now); + + if (!next->tv_sec && !next->tv_nsec) { + /* Not yet initialized. Schedule to get counters one 'period' from 'now': + * Set 'next' to 'now', and the period is added by timespecadd() below. + * (We could retrieve counters immediately -- but at startup counters are then queried even before the + * nft table was created by the maintenance thread. That is not harmful, but it causes an ugly error + * message in the logs. So rather wait one period.) + */ + *next = now; + } + timespecadd(next, &period, next); + if (timespeccmp(next, &now, <)) { + /* The time that has elapsed since last scheduling counter retrieval is already more than the configured + * period. Continue counting the time period from 'now', and ask for counters right now. */ + timespecsub(&now, next, &diff); + LOGP(DNFT, LOGL_NOTICE, "nft-kpi: retrieving counters took %ld.%06ld s longer" + " than the timeout configured in timer hnbgw X34.\n", diff.tv_sec, diff.tv_nsec / 1000); + *next = now; + nft_kpi_get_counters_cb(NULL); + return; + } + + /* next > now, wait for the remaining time. */ + timespecsub(next, &now, &diff); + LOGP(DNFT, LOGL_DEBUG, "nft-kpi: scheduling timer: period is %ld.%06ld s, next occurence in %ld.%06ld s\n", + period.tv_sec, period.tv_nsec / 1000, + diff.tv_sec, diff.tv_nsec / 1000); osmo_timer_setup(&g_hnbgw->nft_kpi.get_counters_timer, nft_kpi_get_counters_cb, NULL); - osmo_timer_schedule(&g_hnbgw->nft_kpi.get_counters_timer, period_s, period_us); + osmo_timer_schedule(&g_hnbgw->nft_kpi.get_counters_timer, diff.tv_sec, diff.tv_nsec / 1000); }

/* from main(), initialize all worker threads and other nft state. */