fix(alarm): fix false sysmem usage alarm

Use information from lc app to trigger sysmem alarm set/clear actions
This commit is contained in:
Zaiming (Stone) Shi 2022-05-24 18:56:28 +02:00
parent 7fc0a84b43
commit 17b9baf509
6 changed files with 235 additions and 81 deletions

View File

@ -58,7 +58,7 @@
, {observer_cli, "1.6.1"} % NOTE: depends on recon 2.5.1 , {observer_cli, "1.6.1"} % NOTE: depends on recon 2.5.1
, {getopt, "1.0.1"} , {getopt, "1.0.1"}
, {snabbkaffe, {git, "https://github.com/kafka4beam/snabbkaffe.git", {tag, "0.15.0"}}} , {snabbkaffe, {git, "https://github.com/kafka4beam/snabbkaffe.git", {tag, "0.15.0"}}}
, {lc, {git, "https://github.com/emqx/lc.git", {tag, "0.2.1"}}} , {lc, {git, "https://github.com/emqx/lc.git", {tag, "0.3.1"}}}
, {mongodb, {git,"https://github.com/emqx/mongodb-erlang", {tag, "v3.0.13"}}} , {mongodb, {git,"https://github.com/emqx/mongodb-erlang", {tag, "v3.0.13"}}}
, {epgsql, {git, "https://github.com/emqx/epgsql.git", {tag, "4.6.0"}}} , {epgsql, {git, "https://github.com/emqx/epgsql.git", {tag, "4.6.0"}}}
]}. ]}.

View File

@ -11,6 +11,9 @@
{load_module,emqx_metrics,brutal_purge,soft_purge,[]}, {load_module,emqx_metrics,brutal_purge,soft_purge,[]},
{apply,{emqx_metrics,assign_auth_stats_from_ets_to_counter,[]}}, {apply,{emqx_metrics,assign_auth_stats_from_ets_to_counter,[]}},
{load_module,emqx_access_control,brutal_purge,soft_purge,[]}, {load_module,emqx_access_control,brutal_purge,soft_purge,[]},
{load_module,emqx_alarm_handler,brutal_purge,soft_purge,[]},
{load_module,emqx_alarm,brutal_purge,soft_purge,[]},
{update,emqx_os_mon,{advanced,[]}},
{load_module,emqx_app,brutal_purge,soft_purge,[]}]}, {load_module,emqx_app,brutal_purge,soft_purge,[]}]},
{"4.3.14", {"4.3.14",
[{load_module,emqx_misc,brutal_purge,soft_purge,[]}, [{load_module,emqx_misc,brutal_purge,soft_purge,[]},
@ -26,6 +29,9 @@
{load_module,emqx_metrics,brutal_purge,soft_purge,[]}, {load_module,emqx_metrics,brutal_purge,soft_purge,[]},
{apply,{emqx_metrics,assign_auth_stats_from_ets_to_counter,[]}}, {apply,{emqx_metrics,assign_auth_stats_from_ets_to_counter,[]}},
{load_module,emqx_access_control,brutal_purge,soft_purge,[]}, {load_module,emqx_access_control,brutal_purge,soft_purge,[]},
{load_module,emqx_alarm_handler,brutal_purge,soft_purge,[]},
{load_module,emqx_alarm,brutal_purge,soft_purge,[]},
{update,emqx_os_mon,{advanced,[]}},
{load_module,emqx_hooks,brutal_purge,soft_purge,[]}]}, {load_module,emqx_hooks,brutal_purge,soft_purge,[]}]},
{"4.3.13", {"4.3.13",
[{load_module,emqx_session,brutal_purge,soft_purge,[]}, [{load_module,emqx_session,brutal_purge,soft_purge,[]},
@ -38,7 +44,6 @@
{load_module,emqx_sys_mon,brutal_purge,soft_purge,[]}, {load_module,emqx_sys_mon,brutal_purge,soft_purge,[]},
{load_module,emqx_pmon,brutal_purge,soft_purge,[]}, {load_module,emqx_pmon,brutal_purge,soft_purge,[]},
{load_module,emqx_banned,brutal_purge,soft_purge,[]}, {load_module,emqx_banned,brutal_purge,soft_purge,[]},
{load_module,emqx_os_mon,brutal_purge,soft_purge,[]},
{load_module,emqx_channel,brutal_purge,soft_purge,[]}, {load_module,emqx_channel,brutal_purge,soft_purge,[]},
{load_module,emqx_sys,brutal_purge,soft_purge,[]}, {load_module,emqx_sys,brutal_purge,soft_purge,[]},
{load_module,emqx_app,brutal_purge,soft_purge,[]}, {load_module,emqx_app,brutal_purge,soft_purge,[]},
@ -48,6 +53,9 @@
{load_module,emqx_access_control,brutal_purge,soft_purge,[]}, {load_module,emqx_access_control,brutal_purge,soft_purge,[]},
{load_module,emqx_cm,brutal_purge,soft_purge,[]}, {load_module,emqx_cm,brutal_purge,soft_purge,[]},
{load_module,emqx_misc,brutal_purge,soft_purge,[]}, {load_module,emqx_misc,brutal_purge,soft_purge,[]},
{load_module,emqx_alarm_handler,brutal_purge,soft_purge,[]},
{load_module,emqx_alarm,brutal_purge,soft_purge,[]},
{update,emqx_os_mon,{advanced,[]}},
{load_module,emqx_connection,brutal_purge,soft_purge,[]}]}, {load_module,emqx_connection,brutal_purge,soft_purge,[]}]},
{"4.3.12", {"4.3.12",
[{load_module,emqx_access_rule,brutal_purge,soft_purge,[]}, [{load_module,emqx_access_rule,brutal_purge,soft_purge,[]},
@ -71,13 +79,15 @@
{load_module,emqx_channel,brutal_purge,soft_purge,[]}, {load_module,emqx_channel,brutal_purge,soft_purge,[]},
{load_module,emqx_session,brutal_purge,soft_purge,[]}, {load_module,emqx_session,brutal_purge,soft_purge,[]},
{load_module,emqx_alarm,brutal_purge,soft_purge,[]}, {load_module,emqx_alarm,brutal_purge,soft_purge,[]},
{load_module,emqx_os_mon,brutal_purge,soft_purge,[]}, {load_module,emqx_alarm_handler,brutal_purge,soft_purge,[]},
{update,emqx_os_mon,{advanced,[]}},
{load_module,emqx,brutal_purge,soft_purge,[]}, {load_module,emqx,brutal_purge,soft_purge,[]},
{load_module,emqx_app,brutal_purge,soft_purge,[]}, {load_module,emqx_app,brutal_purge,soft_purge,[]},
{load_module,emqx_message,brutal_purge,soft_purge,[]}, {load_module,emqx_message,brutal_purge,soft_purge,[]},
{load_module,emqx_limiter,brutal_purge,soft_purge,[]}]}, {load_module,emqx_limiter,brutal_purge,soft_purge,[]}]},
{"4.3.11", {"4.3.11",
[{load_module,emqx_access_rule,brutal_purge,soft_purge,[]}, [{load_module,emqx_alarm_handler,brutal_purge,soft_purge,[]},
{load_module,emqx_access_rule,brutal_purge,soft_purge,[]},
{load_module,emqx_shared_sub,brutal_purge,soft_purge,[]}, {load_module,emqx_shared_sub,brutal_purge,soft_purge,[]},
{load_module,emqx_hooks,brutal_purge,soft_purge,[]}, {load_module,emqx_hooks,brutal_purge,soft_purge,[]},
{load_module,emqx_plugins,brutal_purge,soft_purge,[]}, {load_module,emqx_plugins,brutal_purge,soft_purge,[]},
@ -99,14 +109,15 @@
{load_module,emqx_sys_mon,brutal_purge,soft_purge,[]}, {load_module,emqx_sys_mon,brutal_purge,soft_purge,[]},
{load_module,emqx_frame,brutal_purge,soft_purge,[]}, {load_module,emqx_frame,brutal_purge,soft_purge,[]},
{load_module,emqx_alarm,brutal_purge,soft_purge,[]}, {load_module,emqx_alarm,brutal_purge,soft_purge,[]},
{load_module,emqx_os_mon,brutal_purge,soft_purge,[]}, {update,emqx_os_mon,{advanced,[]}},
{load_module,emqx,brutal_purge,soft_purge,[]}, {load_module,emqx,brutal_purge,soft_purge,[]},
{load_module,emqx_app,brutal_purge,soft_purge,[]}, {load_module,emqx_app,brutal_purge,soft_purge,[]},
{load_module,emqx_http_lib,brutal_purge,soft_purge,[]}, {load_module,emqx_http_lib,brutal_purge,soft_purge,[]},
{load_module,emqx_message,brutal_purge,soft_purge,[]}, {load_module,emqx_message,brutal_purge,soft_purge,[]},
{load_module,emqx_limiter,brutal_purge,soft_purge,[]}]}, {load_module,emqx_limiter,brutal_purge,soft_purge,[]}]},
{"4.3.10", {"4.3.10",
[{load_module,emqx_access_rule,brutal_purge,soft_purge,[]}, [{load_module,emqx_alarm_handler,brutal_purge,soft_purge,[]},
{load_module,emqx_access_rule,brutal_purge,soft_purge,[]},
{load_module,emqx_shared_sub,brutal_purge,soft_purge,[]}, {load_module,emqx_shared_sub,brutal_purge,soft_purge,[]},
{load_module,emqx_hooks,brutal_purge,soft_purge,[]}, {load_module,emqx_hooks,brutal_purge,soft_purge,[]},
{load_module,emqx_plugins,brutal_purge,soft_purge,[]}, {load_module,emqx_plugins,brutal_purge,soft_purge,[]},
@ -129,13 +140,14 @@
{load_module,emqx_app,brutal_purge,soft_purge,[]}, {load_module,emqx_app,brutal_purge,soft_purge,[]},
{load_module,emqx_frame,brutal_purge,soft_purge,[]}, {load_module,emqx_frame,brutal_purge,soft_purge,[]},
{load_module,emqx_alarm,brutal_purge,soft_purge,[]}, {load_module,emqx_alarm,brutal_purge,soft_purge,[]},
{load_module,emqx_os_mon,brutal_purge,soft_purge,[]}, {update,emqx_os_mon,{advanced,[]}},
{load_module,emqx,brutal_purge,soft_purge,[]}, {load_module,emqx,brutal_purge,soft_purge,[]},
{load_module,emqx_connection,brutal_purge,soft_purge,[]}, {load_module,emqx_connection,brutal_purge,soft_purge,[]},
{load_module,emqx_message,brutal_purge,soft_purge,[]}, {load_module,emqx_message,brutal_purge,soft_purge,[]},
{load_module,emqx_limiter,brutal_purge,soft_purge,[]}]}, {load_module,emqx_limiter,brutal_purge,soft_purge,[]}]},
{"4.3.9", {"4.3.9",
[{load_module,emqx_access_rule,brutal_purge,soft_purge,[]}, [{load_module,emqx_alarm_handler,brutal_purge,soft_purge,[]},
{load_module,emqx_access_rule,brutal_purge,soft_purge,[]},
{load_module,emqx_shared_sub,brutal_purge,soft_purge,[]}, {load_module,emqx_shared_sub,brutal_purge,soft_purge,[]},
{load_module,emqx_hooks,brutal_purge,soft_purge,[]}, {load_module,emqx_hooks,brutal_purge,soft_purge,[]},
{load_module,emqx_plugins,brutal_purge,soft_purge,[]}, {load_module,emqx_plugins,brutal_purge,soft_purge,[]},
@ -163,12 +175,13 @@
{load_module,emqx_rpc,brutal_purge,soft_purge,[]}, {load_module,emqx_rpc,brutal_purge,soft_purge,[]},
{load_module,emqx_alarm,brutal_purge,soft_purge,[]}, {load_module,emqx_alarm,brutal_purge,soft_purge,[]},
{load_module,emqx,brutal_purge,soft_purge,[]}, {load_module,emqx,brutal_purge,soft_purge,[]},
{load_module,emqx_os_mon,brutal_purge,soft_purge,[]}, {update,emqx_os_mon,{advanced,[]}},
{load_module,emqx_app,brutal_purge,soft_purge,[]}, {load_module,emqx_app,brutal_purge,soft_purge,[]},
{load_module,emqx_message,brutal_purge,soft_purge,[]}, {load_module,emqx_message,brutal_purge,soft_purge,[]},
{load_module,emqx_limiter,brutal_purge,soft_purge,[]}]}, {load_module,emqx_limiter,brutal_purge,soft_purge,[]}]},
{"4.3.8", {"4.3.8",
[{load_module,emqx_access_rule,brutal_purge,soft_purge,[]}, [{load_module,emqx_alarm_handler,brutal_purge,soft_purge,[]},
{load_module,emqx_access_rule,brutal_purge,soft_purge,[]},
{load_module,emqx_shared_sub,brutal_purge,soft_purge,[]}, {load_module,emqx_shared_sub,brutal_purge,soft_purge,[]},
{load_module,emqx_hooks,brutal_purge,soft_purge,[]}, {load_module,emqx_hooks,brutal_purge,soft_purge,[]},
{load_module,emqx_plugins,brutal_purge,soft_purge,[]}, {load_module,emqx_plugins,brutal_purge,soft_purge,[]},
@ -195,7 +208,7 @@
{load_module,emqx_frame,brutal_purge,soft_purge,[]}, {load_module,emqx_frame,brutal_purge,soft_purge,[]},
{load_module,emqx_rpc,brutal_purge,soft_purge,[]}, {load_module,emqx_rpc,brutal_purge,soft_purge,[]},
{load_module,emqx_alarm,brutal_purge,soft_purge,[]}, {load_module,emqx_alarm,brutal_purge,soft_purge,[]},
{load_module,emqx_os_mon,brutal_purge,soft_purge,[]}, {update,emqx_os_mon,{advanced,[]}},
{load_module,emqx,brutal_purge,soft_purge,[]}, {load_module,emqx,brutal_purge,soft_purge,[]},
{load_module,emqx_app,brutal_purge,soft_purge,[]}, {load_module,emqx_app,brutal_purge,soft_purge,[]},
{load_module,emqx_message,brutal_purge,soft_purge,[]}, {load_module,emqx_message,brutal_purge,soft_purge,[]},
@ -229,7 +242,7 @@
{load_module,emqx_frame,brutal_purge,soft_purge,[]}, {load_module,emqx_frame,brutal_purge,soft_purge,[]},
{load_module,emqx_rpc,brutal_purge,soft_purge,[]}, {load_module,emqx_rpc,brutal_purge,soft_purge,[]},
{load_module,emqx_alarm,brutal_purge,soft_purge,[]}, {load_module,emqx_alarm,brutal_purge,soft_purge,[]},
{load_module,emqx_os_mon,brutal_purge,soft_purge,[]}, {update,emqx_os_mon,{advanced,[]}},
{load_module,emqx,brutal_purge,soft_purge,[]}, {load_module,emqx,brutal_purge,soft_purge,[]},
{load_module,emqx_app,brutal_purge,soft_purge,[]}, {load_module,emqx_app,brutal_purge,soft_purge,[]},
{load_module,emqx_message,brutal_purge,soft_purge,[]}, {load_module,emqx_message,brutal_purge,soft_purge,[]},
@ -263,7 +276,7 @@
{load_module,emqx_frame,brutal_purge,soft_purge,[]}, {load_module,emqx_frame,brutal_purge,soft_purge,[]},
{load_module,emqx_rpc,brutal_purge,soft_purge,[]}, {load_module,emqx_rpc,brutal_purge,soft_purge,[]},
{load_module,emqx_alarm,brutal_purge,soft_purge,[]}, {load_module,emqx_alarm,brutal_purge,soft_purge,[]},
{load_module,emqx_os_mon,brutal_purge,soft_purge,[]}, {update,emqx_os_mon,{advanced,[]}},
{load_module,emqx,brutal_purge,soft_purge,[]}, {load_module,emqx,brutal_purge,soft_purge,[]},
{load_module,emqx_app,brutal_purge,soft_purge,[]}, {load_module,emqx_app,brutal_purge,soft_purge,[]},
{load_module,emqx_message,brutal_purge,soft_purge,[]}, {load_module,emqx_message,brutal_purge,soft_purge,[]},
@ -297,7 +310,7 @@
{load_module,emqx_frame,brutal_purge,soft_purge,[]}, {load_module,emqx_frame,brutal_purge,soft_purge,[]},
{load_module,emqx_rpc,brutal_purge,soft_purge,[]}, {load_module,emqx_rpc,brutal_purge,soft_purge,[]},
{load_module,emqx_alarm,brutal_purge,soft_purge,[]}, {load_module,emqx_alarm,brutal_purge,soft_purge,[]},
{load_module,emqx_os_mon,brutal_purge,soft_purge,[]}, {update,emqx_os_mon,{advanced,[]}},
{load_module,emqx,brutal_purge,soft_purge,[]}, {load_module,emqx,brutal_purge,soft_purge,[]},
{load_module,emqx_app,brutal_purge,soft_purge,[]}, {load_module,emqx_app,brutal_purge,soft_purge,[]},
{load_module,emqx_message,brutal_purge,soft_purge,[]}, {load_module,emqx_message,brutal_purge,soft_purge,[]},
@ -331,7 +344,7 @@
{load_module,emqx_frame,brutal_purge,soft_purge,[]}, {load_module,emqx_frame,brutal_purge,soft_purge,[]},
{load_module,emqx_rpc,brutal_purge,soft_purge,[]}, {load_module,emqx_rpc,brutal_purge,soft_purge,[]},
{load_module,emqx_alarm,brutal_purge,soft_purge,[]}, {load_module,emqx_alarm,brutal_purge,soft_purge,[]},
{load_module,emqx_os_mon,brutal_purge,soft_purge,[]}, {update,emqx_os_mon,{advanced,[]}},
{load_module,emqx,brutal_purge,soft_purge,[]}, {load_module,emqx,brutal_purge,soft_purge,[]},
{load_module,emqx_app,brutal_purge,soft_purge,[]}, {load_module,emqx_app,brutal_purge,soft_purge,[]},
{load_module,emqx_message,brutal_purge,soft_purge,[]}, {load_module,emqx_message,brutal_purge,soft_purge,[]},
@ -366,7 +379,7 @@
{load_module,emqx_frame,brutal_purge,soft_purge,[]}, {load_module,emqx_frame,brutal_purge,soft_purge,[]},
{load_module,emqx_rpc,brutal_purge,soft_purge,[]}, {load_module,emqx_rpc,brutal_purge,soft_purge,[]},
{load_module,emqx_alarm,brutal_purge,soft_purge,[]}, {load_module,emqx_alarm,brutal_purge,soft_purge,[]},
{load_module,emqx_os_mon,brutal_purge,soft_purge,[]}, {update,emqx_os_mon,{advanced,[]}},
{load_module,emqx,brutal_purge,soft_purge,[]}, {load_module,emqx,brutal_purge,soft_purge,[]},
{load_module,emqx_app,brutal_purge,soft_purge,[]}, {load_module,emqx_app,brutal_purge,soft_purge,[]},
{load_module,emqx_message,brutal_purge,soft_purge,[]}, {load_module,emqx_message,brutal_purge,soft_purge,[]},
@ -402,7 +415,7 @@
{load_module,emqx_rpc,brutal_purge,soft_purge,[]}, {load_module,emqx_rpc,brutal_purge,soft_purge,[]},
{load_module,emqx_alarm,brutal_purge,soft_purge,[]}, {load_module,emqx_alarm,brutal_purge,soft_purge,[]},
{load_module,emqx,brutal_purge,soft_purge,[]}, {load_module,emqx,brutal_purge,soft_purge,[]},
{load_module,emqx_os_mon,brutal_purge,soft_purge,[]}, {update,emqx_os_mon,{advanced,[]}},
{load_module,emqx_app,brutal_purge,soft_purge,[]}, {load_module,emqx_app,brutal_purge,soft_purge,[]},
{load_module,emqx_message,brutal_purge,soft_purge,[]}, {load_module,emqx_message,brutal_purge,soft_purge,[]},
{load_module,emqx_limiter,brutal_purge,soft_purge,[]}]}, {load_module,emqx_limiter,brutal_purge,soft_purge,[]}]},
@ -439,7 +452,7 @@
{load_module,emqx_mqueue,brutal_purge,soft_purge,[]}, {load_module,emqx_mqueue,brutal_purge,soft_purge,[]},
{load_module,emqx_rpc,brutal_purge,soft_purge,[]}, {load_module,emqx_rpc,brutal_purge,soft_purge,[]},
{load_module,emqx_alarm,brutal_purge,soft_purge,[]}, {load_module,emqx_alarm,brutal_purge,soft_purge,[]},
{load_module,emqx_os_mon,brutal_purge,soft_purge,[]}, {update,emqx_os_mon,{advanced,[]}},
{load_module,emqx,brutal_purge,soft_purge,[]}, {load_module,emqx,brutal_purge,soft_purge,[]},
{load_module,emqx_app,brutal_purge,soft_purge,[]}, {load_module,emqx_app,brutal_purge,soft_purge,[]},
{load_module,emqx_message,brutal_purge,soft_purge,[]}, {load_module,emqx_message,brutal_purge,soft_purge,[]},
@ -480,14 +493,16 @@
{load_module,emqx_ctl,brutal_purge,soft_purge,[]}, {load_module,emqx_ctl,brutal_purge,soft_purge,[]},
{load_module,emqx_rpc,brutal_purge,soft_purge,[]}, {load_module,emqx_rpc,brutal_purge,soft_purge,[]},
{load_module,emqx_alarm,brutal_purge,soft_purge,[]}, {load_module,emqx_alarm,brutal_purge,soft_purge,[]},
{load_module,emqx_os_mon,brutal_purge,soft_purge,[]}, {update,emqx_os_mon,{advanced,[]}},
{load_module,emqx,brutal_purge,soft_purge,[]}, {load_module,emqx,brutal_purge,soft_purge,[]},
{load_module,emqx_app,brutal_purge,soft_purge,[]}, {load_module,emqx_app,brutal_purge,soft_purge,[]},
{load_module,emqx_message,brutal_purge,soft_purge,[]}, {load_module,emqx_message,brutal_purge,soft_purge,[]},
{load_module,emqx_limiter,brutal_purge,soft_purge,[]}]}, {load_module,emqx_limiter,brutal_purge,soft_purge,[]}]},
{<<".*">>,[]}], {<<".*">>,[]}],
[{"4.3.15", [{"4.3.15",
[{load_module,emqx_misc,brutal_purge,soft_purge,[]}, [{load_module,emqx_alarm,brutal_purge,soft_purge,[]},
{load_module,emqx_alarm_handler,brutal_purge,soft_purge,[]},
{load_module,emqx_misc,brutal_purge,soft_purge,[]},
{load_module,emqx_session,brutal_purge,soft_purge,[]}, {load_module,emqx_session,brutal_purge,soft_purge,[]},
{load_module,emqx_shared_sub,brutal_purge,soft_purge,[]}, {load_module,emqx_shared_sub,brutal_purge,soft_purge,[]},
{load_module,emqx_channel,brutal_purge,soft_purge,[]}, {load_module,emqx_channel,brutal_purge,soft_purge,[]},
@ -495,9 +510,12 @@
{load_module,emqx_frame,brutal_purge,soft_purge,[]}, {load_module,emqx_frame,brutal_purge,soft_purge,[]},
{load_module,emqx_access_rule,brutal_purge,soft_purge,[]}, {load_module,emqx_access_rule,brutal_purge,soft_purge,[]},
{load_module,emqx_access_control,brutal_purge,soft_purge,[]}, {load_module,emqx_access_control,brutal_purge,soft_purge,[]},
{load_module,emqx_os_mon,brutal_purge,soft_purge,[]},
{load_module,emqx_app,brutal_purge,soft_purge,[]}]}, {load_module,emqx_app,brutal_purge,soft_purge,[]}]},
{"4.3.14", {"4.3.14",
[{load_module,emqx_misc,brutal_purge,soft_purge,[]}, [{load_module,emqx_alarm,brutal_purge,soft_purge,[]},
{load_module,emqx_alarm_handler,brutal_purge,soft_purge,[]},
{load_module,emqx_misc,brutal_purge,soft_purge,[]},
{load_module,emqx_session,brutal_purge,soft_purge,[]}, {load_module,emqx_session,brutal_purge,soft_purge,[]},
{load_module,emqx_access_rule,brutal_purge,soft_purge,[]}, {load_module,emqx_access_rule,brutal_purge,soft_purge,[]},
{load_module,emqx,brutal_purge,soft_purge,[]}, {load_module,emqx,brutal_purge,soft_purge,[]},
@ -509,9 +527,12 @@
{load_module,emqx_shared_sub,brutal_purge,soft_purge,[]}, {load_module,emqx_shared_sub,brutal_purge,soft_purge,[]},
{load_module,emqx_frame,brutal_purge,soft_purge,[]}, {load_module,emqx_frame,brutal_purge,soft_purge,[]},
{load_module,emqx_app,brutal_purge,soft_purge,[]}, {load_module,emqx_app,brutal_purge,soft_purge,[]},
{load_module,emqx_os_mon,brutal_purge,soft_purge,[]},
{load_module,emqx_hooks,brutal_purge,soft_purge,[]}]}, {load_module,emqx_hooks,brutal_purge,soft_purge,[]}]},
{"4.3.13", {"4.3.13",
[{load_module,emqx_session,brutal_purge,soft_purge,[]}, [{load_module,emqx_alarm,brutal_purge,soft_purge,[]},
{load_module,emqx_alarm_handler,brutal_purge,soft_purge,[]},
{load_module,emqx_session,brutal_purge,soft_purge,[]},
{load_module,emqx_access_rule,brutal_purge,soft_purge,[]}, {load_module,emqx_access_rule,brutal_purge,soft_purge,[]},
{load_module,emqx_shared_sub,brutal_purge,soft_purge,[]}, {load_module,emqx_shared_sub,brutal_purge,soft_purge,[]},
{load_module,emqx_hooks,brutal_purge,soft_purge,[]}, {load_module,emqx_hooks,brutal_purge,soft_purge,[]},
@ -532,7 +553,8 @@
{load_module,emqx_cm,brutal_purge,soft_purge,[]}, {load_module,emqx_cm,brutal_purge,soft_purge,[]},
{load_module,emqx_connection,brutal_purge,soft_purge,[]}]}, {load_module,emqx_connection,brutal_purge,soft_purge,[]}]},
{"4.3.12", {"4.3.12",
[{load_module,emqx_access_rule,brutal_purge,soft_purge,[]}, [{load_module,emqx_alarm_handler,brutal_purge,soft_purge,[]},
{load_module,emqx_access_rule,brutal_purge,soft_purge,[]},
{load_module,emqx_shared_sub,brutal_purge,soft_purge,[]}, {load_module,emqx_shared_sub,brutal_purge,soft_purge,[]},
{load_module,emqx_hooks,brutal_purge,soft_purge,[]}, {load_module,emqx_hooks,brutal_purge,soft_purge,[]},
{load_module,emqx_plugins,brutal_purge,soft_purge,[]}, {load_module,emqx_plugins,brutal_purge,soft_purge,[]},
@ -557,7 +579,8 @@
{load_module,emqx_message,brutal_purge,soft_purge,[]}, {load_module,emqx_message,brutal_purge,soft_purge,[]},
{load_module,emqx_limiter,brutal_purge,soft_purge,[]}]}, {load_module,emqx_limiter,brutal_purge,soft_purge,[]}]},
{"4.3.11", {"4.3.11",
[{load_module,emqx_access_rule,brutal_purge,soft_purge,[]}, [{load_module,emqx_alarm_handler,brutal_purge,soft_purge,[]},
{load_module,emqx_access_rule,brutal_purge,soft_purge,[]},
{load_module,emqx_shared_sub,brutal_purge,soft_purge,[]}, {load_module,emqx_shared_sub,brutal_purge,soft_purge,[]},
{load_module,emqx_hooks,brutal_purge,soft_purge,[]}, {load_module,emqx_hooks,brutal_purge,soft_purge,[]},
{load_module,emqx_plugins,brutal_purge,soft_purge,[]}, {load_module,emqx_plugins,brutal_purge,soft_purge,[]},
@ -584,7 +607,8 @@
{load_module,emqx_message,brutal_purge,soft_purge,[]}, {load_module,emqx_message,brutal_purge,soft_purge,[]},
{load_module,emqx_limiter,brutal_purge,soft_purge,[]}]}, {load_module,emqx_limiter,brutal_purge,soft_purge,[]}]},
{"4.3.10", {"4.3.10",
[{load_module,emqx_access_rule,brutal_purge,soft_purge,[]}, [{load_module,emqx_alarm_handler,brutal_purge,soft_purge,[]},
{load_module,emqx_access_rule,brutal_purge,soft_purge,[]},
{load_module,emqx_shared_sub,brutal_purge,soft_purge,[]}, {load_module,emqx_shared_sub,brutal_purge,soft_purge,[]},
{load_module,emqx_hooks,brutal_purge,soft_purge,[]}, {load_module,emqx_hooks,brutal_purge,soft_purge,[]},
{load_module,emqx_plugins,brutal_purge,soft_purge,[]}, {load_module,emqx_plugins,brutal_purge,soft_purge,[]},
@ -611,7 +635,8 @@
{load_module,emqx_message,brutal_purge,soft_purge,[]}, {load_module,emqx_message,brutal_purge,soft_purge,[]},
{load_module,emqx_limiter,brutal_purge,soft_purge,[]}]}, {load_module,emqx_limiter,brutal_purge,soft_purge,[]}]},
{"4.3.9", {"4.3.9",
[{load_module,emqx_access_rule,brutal_purge,soft_purge,[]}, [{load_module,emqx_alarm_handler,brutal_purge,soft_purge,[]},
{load_module,emqx_access_rule,brutal_purge,soft_purge,[]},
{load_module,emqx_shared_sub,brutal_purge,soft_purge,[]}, {load_module,emqx_shared_sub,brutal_purge,soft_purge,[]},
{load_module,emqx_hooks,brutal_purge,soft_purge,[]}, {load_module,emqx_hooks,brutal_purge,soft_purge,[]},
{load_module,emqx_plugins,brutal_purge,soft_purge,[]}, {load_module,emqx_plugins,brutal_purge,soft_purge,[]},
@ -642,7 +667,8 @@
{load_module,emqx_message,brutal_purge,soft_purge,[]}, {load_module,emqx_message,brutal_purge,soft_purge,[]},
{load_module,emqx_limiter,brutal_purge,soft_purge,[]}]}, {load_module,emqx_limiter,brutal_purge,soft_purge,[]}]},
{"4.3.8", {"4.3.8",
[{load_module,emqx_access_rule,brutal_purge,soft_purge,[]}, [{load_module,emqx_alarm_handler,brutal_purge,soft_purge,[]},
{load_module,emqx_access_rule,brutal_purge,soft_purge,[]},
{load_module,emqx_shared_sub,brutal_purge,soft_purge,[]}, {load_module,emqx_shared_sub,brutal_purge,soft_purge,[]},
{load_module,emqx_hooks,brutal_purge,soft_purge,[]}, {load_module,emqx_hooks,brutal_purge,soft_purge,[]},
{load_module,emqx_plugins,brutal_purge,soft_purge,[]}, {load_module,emqx_plugins,brutal_purge,soft_purge,[]},

View File

@ -39,6 +39,8 @@
, deactivate/1 , deactivate/1
, deactivate/2 , deactivate/2
, delete_all_deactivated_alarms/0 , delete_all_deactivated_alarms/0
, ensure_deactivated/1
, ensure_deactivated/2
, get_alarms/0 , get_alarms/0
, get_alarms/1 , get_alarms/1
]). ]).
@ -132,6 +134,24 @@ activate(Name) ->
activate(Name, Details) -> activate(Name, Details) ->
gen_server:call(?MODULE, {activate_alarm, Name, Details}). gen_server:call(?MODULE, {activate_alarm, Name, Details}).
-spec ensure_deactivated(binary() | atom()) -> ok.
ensure_deactivated(Name) ->
ensure_deactivated(Name, no_details).
-spec ensure_deactivated(binary() | atom(), atom() | map()) -> ok.
ensure_deactivated(Name, Data) ->
%% this duplicates the dirty read in handle_call,
%% intention is to avoid making gen_server calls when there is no alarm
case mnesia:dirty_read(?ACTIVATED_ALARM, Name) of
[] ->
ok;
_ ->
case deactivate(Name, Data) of
{error, not_found} -> ok;
Other -> Other
end
end.
deactivate(Name) -> deactivate(Name) ->
gen_server:call(?MODULE, {deactivate_alarm, Name, no_details}). gen_server:call(?MODULE, {deactivate_alarm, Name, no_details}).

View File

@ -56,21 +56,12 @@ init({_Args, {alarm_handler, _ExistingAlarms}}) ->
init(_) -> init(_) ->
{ok, []}. {ok, []}.
handle_event({set_alarm, {system_memory_high_watermark, []}}, State) ->
emqx_alarm:activate(high_system_memory_usage, #{high_watermark => emqx_os_mon:get_sysmem_high_watermark()}),
{ok, State};
handle_event({set_alarm, {process_memory_high_watermark, Pid}}, State) -> handle_event({set_alarm, {process_memory_high_watermark, Pid}}, State) ->
emqx_alarm:activate(high_process_memory_usage, #{pid => list_to_binary(pid_to_list(Pid)), emqx_alarm:activate(high_process_memory_usage, #{pid => list_to_binary(pid_to_list(Pid)),
high_watermark => emqx_os_mon:get_procmem_high_watermark()}), high_watermark => emqx_os_mon:get_procmem_high_watermark()}),
{ok, State}; {ok, State};
handle_event({clear_alarm, system_memory_high_watermark}, State) ->
emqx_alarm:deactivate(high_system_memory_usage),
{ok, State};
handle_event({clear_alarm, process_memory_high_watermark}, State) -> handle_event({clear_alarm, process_memory_high_watermark}, State) ->
emqx_alarm:deactivate(high_process_memory_usage), emqx_alarm:ensure_deactivate(high_process_memory_usage),
{ok, State}; {ok, State};
handle_event(_, State) -> handle_event(_, State) ->

View File

@ -78,32 +78,30 @@ set_cpu_low_watermark(Float) ->
call({set_cpu_low_watermark, Float}). call({set_cpu_low_watermark, Float}).
get_mem_check_interval() -> get_mem_check_interval() ->
memsup:get_check_interval() div 1000. call(?FUNCTION_NAME).
set_mem_check_interval(Seconds) when Seconds < 60 ->
memsup:set_check_interval(1);
set_mem_check_interval(Seconds) -> set_mem_check_interval(Seconds) ->
memsup:set_check_interval(Seconds div 60). call({?FUNCTION_NAME, Seconds}).
get_sysmem_high_watermark() -> get_sysmem_high_watermark() ->
memsup:get_sysmem_high_watermark(). call(?FUNCTION_NAME).
set_sysmem_high_watermark(Float) -> set_sysmem_high_watermark(HW) ->
V = Float/100, V = resolve_watermark(HW),
case load_ctl:get_config() of case load_ctl:get_config() of
#{ ?MEM_MON_F0 := true } = OldLC -> #{ ?MEM_MON_F0 := true } = OldLC ->
ok = load_ctl:put_config(OldLC#{ ?MEM_MON_F0 => true ok = load_ctl:put_config(OldLC#{ ?MEM_MON_F0 => true
, ?MEM_MON_F1 => V}); , ?MEM_MON_F1 => V / 100});
_ -> _ ->
skip skip
end, end,
memsup:set_sysmem_high_watermark(V). gen_server:call(?OS_MON, {?FUNCTION_NAME, V}, infinity).
get_procmem_high_watermark() -> get_procmem_high_watermark() ->
memsup:get_procmem_high_watermark(). memsup:get_procmem_high_watermark().
set_procmem_high_watermark(Float) -> set_procmem_high_watermark(HW) ->
memsup:set_procmem_high_watermark(Float / 100). memsup:set_procmem_high_watermark(resolve_watermark(HW) / 100).
call(Req) -> call(Req) ->
gen_server:call(?OS_MON, Req, infinity). gen_server:call(?OS_MON, Req, infinity).
@ -113,16 +111,38 @@ call(Req) ->
%%-------------------------------------------------------------------- %%--------------------------------------------------------------------
init([Opts]) -> init([Opts]) ->
set_mem_check_interval(proplists:get_value(mem_check_interval, Opts)), process_flag(trap_exit, true),
SysHW = proplists:get_value(sysmem_high_watermark, Opts), %% make sure memsup will not emit system memory alarms
set_sysmem_high_watermark(SysHW), memsup:set_sysmem_high_watermark(1),
set_procmem_high_watermark(proplists:get_value(procmem_high_watermark, Opts)), set_procmem_high_watermark(proplists:get_value(procmem_high_watermark, Opts)),
ensure_system_memory_alarm(SysHW), MemCehckInterval = do_resolve_mem_check_interval(proplists:get_value(mem_check_interval, Opts)),
{ok, ensure_check_timer(#{cpu_high_watermark => proplists:get_value(cpu_high_watermark, Opts), SysHW = resolve_watermark(proplists:get_value(sysmem_high_watermark, Opts)),
St = ensure_check_timer(#{cpu_high_watermark => proplists:get_value(cpu_high_watermark, Opts),
cpu_low_watermark => proplists:get_value(cpu_low_watermark, Opts), cpu_low_watermark => proplists:get_value(cpu_low_watermark, Opts),
cpu_check_interval => proplists:get_value(cpu_check_interval, Opts), cpu_check_interval => proplists:get_value(cpu_check_interval, Opts),
timer => undefined})}. sysmem_high_watermark => SysHW,
mem_check_interval => MemCehckInterval,
timer => undefined}),
ok = do_set_mem_check_interval(MemCehckInterval),
%% update immediately after start/restart
ok = update_mem_alarm_status(SysHW),
{ok, ensure_mem_check_timer(St)}.
handle_call(get_sysmem_high_watermark, _From, State) ->
#{sysmem_high_watermark := SysHW} = State,
{reply, maybe_round(SysHW), State};
handle_call(get_mem_check_interval, _From, State) ->
#{mem_check_interval := Interval} = State,
{reply, Interval, State};
handle_call({set_sysmem_high_watermark, SysHW}, _From, State) ->
%% update immediately after start/restart
ok = update_mem_alarm_status(SysHW),
{reply, ok, State#{sysmem_high_watermark => SysHW}};
handle_call({set_mem_check_interval, Seconds0}, _From, State) ->
Seconds = do_resolve_mem_check_interval(Seconds0),
ok = do_set_mem_check_interval(Seconds),
%% will start taking effect when the current timer expires
{reply, ok, State#{mem_check_interval => Seconds}};
handle_call(get_cpu_check_interval, _From, State) -> handle_call(get_cpu_check_interval, _From, State) ->
{reply, maps:get(cpu_check_interval, State, undefined), State}; {reply, maps:get(cpu_check_interval, State, undefined), State};
@ -145,6 +165,8 @@ handle_call(Req, _From, State) ->
?LOG(error, "Unexpected call: ~p", [Req]), ?LOG(error, "Unexpected call: ~p", [Req]),
{reply, ignored, State}. {reply, ignored, State}.
handle_cast(upgrade, State) ->
{noreply, State};
handle_cast(Msg, State) -> handle_cast(Msg, State) ->
?LOG(error, "Unexpected cast: ~p", [Msg]), ?LOG(error, "Unexpected cast: ~p", [Msg]),
{noreply, State}. {noreply, State}.
@ -168,16 +190,28 @@ handle_info({timeout, Timer, check}, State = #{timer := Timer,
ensure_check_timer(State) ensure_check_timer(State)
end, end,
{noreply, NState}; {noreply, NState};
handle_info({timeout, Timer, check_mem}, #{mem_check_timer := Timer,
sysmem_high_watermark := SysHW
} = State) ->
ok = update_mem_alarm_status(SysHW),
NState = ensure_mem_check_timer(State#{mem_check_timer := undefined}),
{noreply, NState};
handle_info(Info, State) -> handle_info(Info, State) ->
?LOG(error, "unexpected info: ~p", [Info]), ?LOG(error, "unexpected info: ~p", [Info]),
{noreply, State}. {noreply, State}.
terminate(_Reason, #{timer := Timer}) -> terminate(_Reason, #{timer := Timer} = St) ->
emqx_misc:cancel_timer(maps:get(mem_check_timer, St, undefined)),
emqx_misc:cancel_timer(Timer). emqx_misc:cancel_timer(Timer).
code_change(_OldVsn, State, _Extra) -> code_change(_OldVsn, State, _Extra) ->
{ok, State}. %% NOTE: downgrade is not handled as the extra fields added to State
%% does not affect old version code.
%% The only thing which may slip through is that a started timer
%% will result in a "unexpected info" error log for the old version code
NewState = ensure_mem_check_timer(State),
SysHW = resolve_sysmem_high_watermark(State),
{ok, NewState#{sysmem_high_watermark => SysHW}}.
%%-------------------------------------------------------------------- %%--------------------------------------------------------------------
%% Internal functions %% Internal functions
@ -189,19 +223,93 @@ ensure_check_timer(State = #{cpu_check_interval := Interval}) ->
_ -> State#{timer := emqx_misc:start_timer(timer:seconds(Interval), check)} _ -> State#{timer := emqx_misc:start_timer(timer:seconds(Interval), check)}
end. end.
%% At startup, memsup starts first and checks for memory alarms, ensure_mem_check_timer(#{mem_check_timer := Ref} = State) when is_reference(Ref) ->
%% but emqx_alarm_handler is not yet used instead of alarm_handler, %% timer already started
%% so alarm_handler is used directly for notification (normally emqx_alarm_handler should be used). State;
%%The internal memsup will no longer trigger events that have been alerted, ensure_mem_check_timer(State) ->
%% and there is no exported function to remove the alerted flag, Interval = resolve_mem_check_interval(State),
%% so it can only be checked again at startup. case is_sysmem_check_supported() of
ensure_system_memory_alarm(HW) -> true ->
case erlang:whereis(memsup) of State#{mem_check_timer => emqx_misc:start_timer(timer:seconds(Interval), check_mem),
undefined -> ok; mem_check_interval => Interval
_Pid -> };
{Total, Allocated, _Worst} = memsup:get_memory_data(), false ->
case Total =/= 0 andalso Allocated/Total * 100 > HW of State
true -> emqx_alarm:activate(high_system_memory_usage, #{high_watermark => HW}); end.
false -> ok
end resolve_mem_check_interval(#{mem_check_interval := Seconds}) when is_integer(Seconds) ->
Seconds;
resolve_mem_check_interval(_) ->
%% this only happens when hot-upgrade from older version (< 4.3.14, or < 4.4.4)
try
%% memsup has interval set API using minutes, but returns in milliseconds from get API
IntervalMs = memsup:get_check_interval(),
true = (IntervalMs > 1000),
IntervalMs div 1000
catch
_ : _ ->
%% this is the memsup default
60
end.
is_sysmem_check_supported() ->
%% sorry Mac and Windows, for now
{unix, linux} =:= os:type().
%% we still need to set memsup interval for process (not system) memory check
do_set_mem_check_interval(Seconds) ->
Minutes = Seconds div 60,
_ = memsup:set_check_interval(Minutes),
ok.
%% keep the time unit alignment with memsup, minmum interval is 60 seconds.
do_resolve_mem_check_interval(Seconds) ->
case is_integer(Seconds) andalso Seconds >= 60 of
true -> Seconds;
false -> 60
end.
resolve_sysmem_high_watermark(#{sysmem_high_watermark := SysHW}) -> SysHW;
resolve_sysmem_high_watermark(_) ->
%% sysmem_high_watermark is not found in state map
%% get it from memsup
resolve_watermark(memsup:get_sysmem_high_watermark()).
resolve_watermark(W) when W > 0 andalso W =< 1 ->
W * 100;
resolve_watermark(W) when W > 0 andalso W =< 100 ->
W.
update_mem_alarm_status(SysHW) ->
Usage = current_sysmem_percent(),
case Usage > SysHW of
true ->
_ = emqx_alarm:activate(
high_system_memory_usage,
#{
usage => Usage,
high_watermark => SysHW
}
);
_ ->
ok = emqx_alarm:ensure_deactivated(
high_system_memory_usage,
#{
usage => Usage,
high_watermark => SysHW
}
)
end,
ok.
current_sysmem_percent() ->
Ratio = load_ctl:get_memory_usage(),
erlang:floor(Ratio * 10000) / 100.
maybe_round(X) when is_integer(X) -> X;
maybe_round(X) when is_float(X) ->
R = erlang:round(X),
case erlang:abs(X - R) > 1.0e-6 of
true -> X;
false -> R
end. end.

View File

@ -43,14 +43,23 @@ end_per_suite(_Config) ->
emqx_ct_helpers:stop_apps([]), emqx_ct_helpers:stop_apps([]),
application:stop(os_mon). application:stop(os_mon).
% t_set_mem_check_interval(_) -> t_set_mem_check_interval(_) ->
% error('TODO'). emqx_os_mon:set_mem_check_interval(0),
?assertEqual(60, emqx_os_mon:get_mem_check_interval()),
emqx_os_mon:set_mem_check_interval(61),
?assertEqual(61, emqx_os_mon:get_mem_check_interval()),
ok.
% t_set_sysmem_high_watermark(_) -> t_set_sysmem_high_watermark(_) ->
% error('TODO'). emqx_os_mon:set_sysmem_high_watermark(10),
?assertEqual(10, emqx_os_mon:get_sysmem_high_watermark()),
% t_set_procmem_high_watermark(_) -> emqx_os_mon:set_sysmem_high_watermark(100),
% error('TODO'). ?assertEqual(100, emqx_os_mon:get_sysmem_high_watermark()),
emqx_os_mon:set_sysmem_high_watermark(0.9),
?assertEqual(90, emqx_os_mon:get_sysmem_high_watermark()),
emqx_os_mon:set_sysmem_high_watermark(0.932),
?assertEqual(93.2, emqx_os_mon:get_sysmem_high_watermark()),
ok.
t_api(_) -> t_api(_) ->
?assertEqual(1, emqx_os_mon:get_cpu_check_interval()), ?assertEqual(1, emqx_os_mon:get_cpu_check_interval()),