diff options
Diffstat (limited to 'lib/diameter')
| -rw-r--r-- | lib/diameter/src/base/diameter_service.erl | 112 |
1 files changed, 54 insertions, 58 deletions
diff --git a/lib/diameter/src/base/diameter_service.erl b/lib/diameter/src/base/diameter_service.erl index 72b886bcdd..91e7cbd996 100644 --- a/lib/diameter/src/base/diameter_service.erl +++ b/lib/diameter/src/base/diameter_service.erl @@ -135,7 +135,12 @@ %% service record is used to determine whether or not we need to call %% the process for a pick_peer callback. -%% Record representing a watchdog process. +%% Record representing a watchdog process as implemented by +%% diameter_watchdog. The term "peer" here is historical, made +%% especially confusing by the fact that a peer_ref() in the +%% documentation is the key of a #conn{} record, not a #peer{} record. +%% The name is also unfortunate given the meaning of peer in the +%% Diameter sense. -record(peer, {pid :: match(pid()), type :: match(connect | accept), @@ -147,7 +152,13 @@ conn = false :: match(boolean() | pid())}). %% true at accepted, pid() at connection_up or reopen -%% Record representing a peer_fsm process. +%% Record representing a peer process as implemented by +%% diameter_peer_fsm. The term "conn" is historical. Despite the name +%% here, comments refer to watchdog and peer processes, that are keys +%% in #peer{} and #conn{} records respectively. To add to the +%% confusion, a #request.transport is a peer process = key in a +%% #conn{} record. The actual transport process (that the peer process +%% knows about and that has a transport connection) isn't seen here. -record(conn, {pid :: pid(), apps :: [{0..16#FFFFFFFF, diameter:app_alias()}], %% {Id, Alias} @@ -161,10 +172,9 @@ handler :: match(pid()), %% request process transport :: match(pid()), %% peer process caps :: match(#diameter_caps{}), - app :: match(diameter:app_alias()), %% #diameter_app.alias - dictionary :: match(module()), %% #diameter_app.dictionary - module :: match([module() | list()]), - %% #diameter_app.module + app :: match(diameter:app_alias()),%% #diameter_app.alias + dictionary :: match(module()), %% #diameter_app.dictionary + module :: match([module() | list()]), %% #diameter_app.module filter :: match(diameter:peer_filter()), packet :: match(#diameter_packet{})}). @@ -175,20 +185,6 @@ timeout = ?DEFAULT_TIMEOUT :: 0..16#FFFFFFFF, detach = false :: boolean()}). -%% Since RFC 3588 requires that a Diameter agent not modify End-to-End -%% Identifiers, the possibility of explicitly setting an End-to-End -%% Identifier would be needed to be able to implement an agent in -%% which one side of the communication is not implemented on top of -%% diameter. For example, Diameter being sent or received encapsulated -%% in some other protocol, or even another Diameter stack in a -%% non-Erlang environment. (Not that this is likely to be a normal -%% case.) -%% -%% The implemented solution is not an option but to respect any header -%% values set in a diameter_header record returned from a -%% prepare_request callback. A call to diameter:call/4 can communicate -%% values to the callback using the 'extra' option if so desired. - %%% --------------------------------------------------------------------------- %%% # start(SvcName) %%% --------------------------------------------------------------------------- @@ -252,9 +248,9 @@ info(SvcName, Item) -> %%% # receive_message(TPid, Pkt, MessageData) %%% --------------------------------------------------------------------------- -%% Handle an incoming message in the watchdog process. This used to -%% come through the service process but this avoids that becoming a -%% bottleneck. +%% Handle an incoming Diameter message in the watchdog process. This +%% used to come through the service process but this avoids that +%% becoming a bottleneck. receive_message(TPid, Pkt, T) when is_pid(TPid) -> @@ -543,20 +539,20 @@ transition({connection_up, Pid, T}, S) -> connection_up(Pid, T, S), ok; -%% Peer process has a new connection that will be opened after -%% watchdog exchange. This message was added long after connection_up, -%% to communicate the information as soon as it's available. Leave +%% Watchdog has a new connection that will be opened after DW[RA] +%% exchange. This message was added long after connection_up, to +%% communicate the information as soon as it's available. Leave %% connection_up as is it for now, duplicated information and all. transition({reopen, Pid, T}, S) -> reopen(Pid, T, S), ok; -%% Peer process has left state open. +%% Watchdog has left state OKAY. transition({connection_down, Pid}, S) -> connection_down(Pid, S), ok; -%% Peer process has returned to state open. +%% Watchdog has returned to state OKAY. transition({connection_up, Pid}, S) -> connection_up(Pid, S), ok; @@ -573,7 +569,7 @@ transition({reconnect, Pid}, S) -> %% Watchdog is sending notification of a state transition. Note that %% the connection_up/down messages are pre-date this message and are -%% still used. A 'watchdog' message will follow these and communicate +%% still used. A watchdog message will follow these and communicate %% the same state as was set in handling connection_up/down. transition({watchdog, Pid, {TPid, From, To}}, #state{service_name = SvcName, peerT = PeerT}) -> @@ -583,21 +579,22 @@ transition({watchdog, Pid, {TPid, From, To}}, #state{service_name = SvcName, insert(PeerT, P#peer{op_state = {OS, To}}), send_event(SvcName, {watchdog, Ref, TPid, {From, To}, {T, Opts}}), ok; -%% Death of a peer process results in the removal of it's peer and any -%% associated conn record when 'DOWN' is received (after this) but the -%% states will be {?STATE_UP, ?WD_DOWN} for a short time. (No real -%% problem since ?WD_* is only used in service_info.) We set ?WD_OKAY -%% as a consequence of connection_up since we know a watchdog is -%% coming. We can't set anything at connection_down since we don't -%% know if the subsequent watchdog message will be ?WD_DOWN or -%% ?WD_SUSPECT. We don't (yet) set ?STATE_* as a consequence of a -%% watchdog message since this requires changing some of the matching -%% on ?STATE_*. -%% -%% Death of a conn process results in connection_down followed by -%% watchdog ?WD_DOWN. The latter doesn't result in the conn record -%% being deleted since 'DOWN' from death of its peer doesn't (yet) -%% deal with the record having been removed. +%% Death of a watchdog process (#peer.pid) results in the removal of +%% it's peer and any associated conn record when 'DOWN' is received +%% (after this) but the states will be {?STATE_UP, ?WD_DOWN} for a +%% short time. (No real problem since ?WD_* is only used in +%% service_info.) We set ?WD_OKAY as a consequence of connection_up +%% since we know a watchdog is coming. We can't set anything at +%% connection_down since we don't know if the subsequent watchdog +%% message will be ?WD_DOWN or ?WD_SUSPECT. We don't (yet) set +%% ?STATE_* as a consequence of a watchdog message since this requires +%% changing some of the matching on ?STATE_*. +%% +%% Death of a peer process process (#conn.pid, #peer.conn) results in +%% connection_down followed by watchdog ?WD_DOWN. The latter doesn't +%% result in the conn record being deleted since 'DOWN' from death of +%% its watchdog doesn't (yet) deal with the record having been +%% removed. %% Monitor process has died. Just die with a reason that tells %% diameter_config about the happening. If a cleaner shutdown is @@ -605,13 +602,13 @@ transition({watchdog, Pid, {TPid, From, To}}, #state{service_name = SvcName, transition({'DOWN', MRef, process, _, Reason}, #state{monitor = MRef}) -> {stop, {monitor, Reason}}; -%% Local peer process has died. +%% Local watchdog process has died. transition({'DOWN', _, process, Pid, Reason}, S) when node(Pid) == node() -> peer_down(Pid, Reason, S), ok; -%% Remote service wants to know about shared transports. +%% Remote service wants to know about shared peers. transition({service, Pid}, S) -> share_peers(Pid, S), ok; @@ -944,9 +941,8 @@ start(Ref, Type, Opts, #state{peerT = PeerT, Pid. %% Note that the service record passed into the watchdog is the merged -%% record so that each watchdog (and peer_fsm) may get a different -%% record. This record is what is passed back into application -%% callbacks. +%% record so that each watchdog may get a different record. This +%% record is what is passed back into application callbacks. s(Type, Ref, T) -> case diameter_watchdog:start({Type, Ref}, T) of @@ -997,8 +993,8 @@ accepted(Pid, _TPid, #state{peerT = PeerT} = S) -> #peer{ref = Ref, type = accept = T, conn = false, options = Opts} = P = fetch(PeerT, Pid), - insert(PeerT, P#peer{conn = true}), %% mark replacement transport started - start(Ref, T, Opts, S). %% start new peer + insert(PeerT, P#peer{conn = true}), %% mark replacement as started + start(Ref, T, Opts, S). %% start new watchdog fetch(Tid, Key) -> [T] = ets:lookup(Tid, Key), @@ -1015,7 +1011,7 @@ fetch(Tid, Key) -> %%% # connection_up/3 %%% --------------------------------------------------------------------------- -%% Peer process has reached the open state. +%% Watchdog process has reached state OKAY. connection_up(Pid, {TPid, {Caps, SApps, Pkt}}, #state{peerT = PeerT, connT = ConnT} @@ -1126,7 +1122,7 @@ peer_cb(MFA, Alias) -> %%% # connection_down/2 %%% --------------------------------------------------------------------------- -%% Peer process has transitioned out of the open state. +%% Watchdog has transitioned out of state OKAY. connection_down(Pid, #state{peerT = PeerT, connT = ConnT} @@ -1178,7 +1174,7 @@ down_conn(Id, Alias, TC, {SvcName, Apps}) -> %%% # peer_down/3 %%% --------------------------------------------------------------------------- -%% Peer process has died. +%% Watchdog process has died. peer_down(Pid, Reason, #state{peerT = PeerT} = S) -> P = fetch(PeerT, Pid), @@ -1198,7 +1194,7 @@ closed({shutdown, {close, _TPid, Reason}}, closed(_, _, _) -> ok. -%% The peer has never come up ... +%% The watchdog has never reached OKAY ... peer_down(#peer{conn = B}, _) when is_boolean(B) -> ok; @@ -1231,7 +1227,7 @@ restart(#peer{ref = Ref, started = Time}) -> {Time, {Ref, T, Opts}}; -%% ... or it has: a replacement transport has already been spawned. +%% ... or it has: a replacement has already been spawned. restart(#peer{type = accept}) -> false. @@ -1257,8 +1253,8 @@ default_tc(connect, Opts) -> default_tc(accept, _) -> 0. -%% Bound tc below if the peer was restarted recently to avoid -%% continuous in case of faulty config or other problems. +%% Bound tc below if the watchdog was restarted recently to avoid +%% continuous restarted in case of faulty config or other problems. tc(Time, Tc) -> choose(Tc > ?RESTART_TC orelse timer:now_diff(now(), Time) > 1000*?RESTART_TC, |