diff options
Diffstat (limited to 'lib/kernel/src/logger_h_common.erl')
| -rw-r--r-- | lib/kernel/src/logger_h_common.erl | 16 |
1 files changed, 10 insertions, 6 deletions
diff --git a/lib/kernel/src/logger_h_common.erl b/lib/kernel/src/logger_h_common.erl index 7caad366ae..901c4c0dad 100644 --- a/lib/kernel/src/logger_h_common.erl +++ b/lib/kernel/src/logger_h_common.erl @@ -135,7 +135,8 @@ call_cast_or_drop(Name, Bin) -> _:{timeout,_} -> ?observe(Name,{dropped,1}) end; - drop -> ?observe(Name,{dropped,1}) + drop -> + ?observe(Name,{dropped,1}) catch %% if the ETS table doesn't exist (maybe because of a %% handler restart), we can only drop the request @@ -152,12 +153,15 @@ check_load(State = #{id:=Name, mode := Mode, flush_reqs_qlen := FlushQLen}) -> {_,Mem} = process_info(self(), memory), ?observe(Name,{max_mem,Mem}), - %% make sure the handler process doesn't get scheduled - %% out between the message_queue_len check below and the - %% action that follows (flush or write). {_,QLen} = process_info(self(), message_queue_len), ?observe(Name,{max_qlen,QLen}), - + %% When the handler process gets scheduled in, it's impossible + %% to predict the QLen. We could jump "up" arbitrarily from say + %% async to sync, async to drop, sync to flush, etc. However, when + %% the handler process manages the log requests (without flushing), + %% one after the other, we will move "down" from drop to sync and + %% from sync to async. This way we don't risk getting stuck in + %% drop or sync mode with an empty mailbox. {Mode1,_NewDrops,_NewFlushes} = if QLen >= FlushQLen -> @@ -292,7 +296,7 @@ overload_levels_ok(HandlerConfig) -> TSQL = maps:get(toggle_sync_qlen, HandlerConfig, ?TOGGLE_SYNC_QLEN), DNRQL = maps:get(drop_new_reqs_qlen, HandlerConfig, ?DROP_NEW_REQS_QLEN), FRQL = maps:get(flush_reqs_qlen, HandlerConfig, ?FLUSH_REQS_QLEN), - (TSQL < DNRQL) andalso (DNRQL < FRQL). + (DNRQL > 1) andalso (TSQL =< DNRQL) andalso (DNRQL =< FRQL). error_notify(Term) -> ?internal_log(error, Term). |