[checkmk-commits] 6227 FIX Fix failing termination or config reload

Module: check_mk Branch: master Commit: fc519a21c8a878af8e0175437c2621037d78fdac URL: http://git.mathias-kettner.de/git/?p=check_mk.git;a=commit;h=fc519a21c8a878… Author: Konstantin Büttner <kb(a)mathias-kettner.de> Date: Fri Jul 13 10:49:09 2018 +0200 6227 FIX Fix failing termination or config reload Change-Id: Ifb0b0e0a2aa49cc12c54bea618ccf7f4b6c756e2 --- .werks/6227 | 10 +++++++++ cmk/ec/main.py | 69 +++++++++++++++++++++++++++++----------------------------- 2 files changed, 45 insertions(+), 34 deletions(-) diff --git a/.werks/6227 b/.werks/6227 new file mode 100644 index 0000000..a4ee1a4 --- /dev/null +++ b/.werks/6227 @@ -0,0 +1,10 @@ +Title: Fix failing termination or config reload +Level: 2 +Component: ec +Class: fix +Compatible: compat +Edition: cre +State: unknown +Version: 1.6.0i1 +Date: 1531469606 + diff --git a/cmk/ec/main.py b/cmk/ec/main.py index 93a0386..0a0526e 100644 --- a/cmk/ec/main.py +++ b/cmk/ec/main.py @@ -3183,35 +3183,41 @@ def run_eventd(terminate_main_event, settings, config, lock_configuration, histo while not terminate_main_event.is_set(): try: - # Wait until either housekeeping or retention is due, but at - # maximum 60 seconds. That way changes of the interval from a very - # high to a low value will never require more than 60 seconds - - event_list = [next_housekeeping, next_retention, next_statistics] - if is_replication_slave(config): - event_list.append(next_replication) - - time_left = max(0, min(event_list) - time.time()) - time.sleep(min(time_left, 60)) - - now = time.time() - if now > next_housekeeping: - event_server.do_housekeeping() - next_housekeeping = now + config["housekeeping_interval"] - - if now > next_retention: - with event_status.lock: - event_status.save_status() - next_retention = now + config["retention_interval"] - - if now > next_statistics: - perfcounters.do_statistics() - next_statistics = now + config["statistics_interval"] - - # Beware: replication might be turned on during this loop! - if is_replication_slave(config) and now > next_replication: - replication_pull(settings, config, lock_configuration, perfcounters, event_status, event_server, slave_status, logger) - next_replication = now + config["replication"]["interval"] + try: + # Wait until either housekeeping or retention is due, but at + # maximum 60 seconds. That way changes of the interval from a very + # high to a low value will never require more than 60 seconds + + event_list = [next_housekeeping, next_retention, next_statistics] + if is_replication_slave(config): + event_list.append(next_replication) + + time_left = max(0, min(event_list) - time.time()) + time.sleep(min(time_left, 60)) + + now = time.time() + if now > next_housekeeping: + event_server.do_housekeeping() + next_housekeeping = now + config["housekeeping_interval"] + + if now > next_retention: + with event_status.lock: + event_status.save_status() + next_retention = now + config["retention_interval"] + + if now > next_statistics: + perfcounters.do_statistics() + next_statistics = now + config["statistics_interval"] + + # Beware: replication might be turned on during this loop! + if is_replication_slave(config) and now > next_replication: + replication_pull(settings, config, lock_configuration, perfcounters, event_status, event_server, slave_status, logger) + next_replication = now + config["replication"]["interval"] + except Exception as e: + logger.exception("Exception in main thread:\n%s" % e) + if settings.options.debug: + raise + time.sleep(1) except MKSignalException as e: if e._signum == 1: logger.info("Received SIGHUP - going to reload configuration") @@ -3219,11 +3225,6 @@ def run_eventd(terminate_main_event, settings, config, lock_configuration, histo else: logger.info("Signalled to death by signal %d" % e._signum) terminate(terminate_main_event, event_server, status_server) - except Exception as e: - logger.exception("Exception in main thread:\n%s" % e) - if settings.options.debug: - raise - time.sleep(1) # Now wait for termination of the server threads event_server.join()