Improve design doc

doc/aaa/TACACS+ Server Monitor.md

@@ -13,19 +13,57 @@ There is a timeout for each server, the default value is 5 seconds, this means i
 To improve this issue, SONiC will add a TACACS+ server monitor to change server priority, a server unreachable or slow response will be downgrade.
 
 ### Functional Requirement
-- Monit TACACS+ server unreachable event from COUNTER_DB.
-- Monit TACACS+ server slow response event from COUNTER_DB.
-- Change server priority based unreachable event and slow response event.
-- Not change any other server attribute.
-- Not change any other TACACS+ config.
+- Monit requirement:
+    - Add a new monit unit for TACACS+ server status check.
+    - Check TACACS+ server connection and write network latency data to COUNTER_DB.
+    - Write TACACS+ server unreachable message to syslog.
+    - Write TACACS+ server slow response message to syslog.
+- Hostcfgd requirement:
+    - Change server priority based on TACACS+ network latency data in COUNTER_DB.
+    - Not change any other server attribute.
+    - Not change any other TACACS+ config.
 
 ### Counter DB schema
 #### TACPLUS_SERVER_LATENCY Table schema
 ```
 ; Key
-server_key           = IPAddress  ;  TACACS+ server’s address
+server_key           = String     ;  TACACS+ server’s address and port
 ; Attributes
-latency              = 1*10DIGIT  ; server network latency in MS, -1 for connect to server timeout
+latency              = 1*10DIGIT  ; Server network latency in MS, -1 for connect to server timeout
+```
+
+#### TACPLUS_SERVER_LATENCY Table YANG model
+```yangmodule sonic-tacplus-server-latency {
+    namespace "http://github.com/sonic-net/sonic-tacplus-server-latency";
+    prefix ssys;
+    yang-version 1.1;
+
+    revision 2023-11-01 {
+        description "Initial revision.";
+    }
+
+    container sonic-tacplus-server-latency {
+
+        container TACPLUS_SERVER_LATENCY {
+
+            list TACPLUS_SERVER_LATENCY_LIST {
+                max-elements 8;
+                key "server";
+
+                leaf server {
+                    type string;
+                    description
+                        "TACACS+ server’s address and port";
+                }
+
+                leaf latency {
+                    type uint32;
+                    description "Server network latency in MS";
+                }
+            }
+        }
+    }
+}
 ```
 
 ### Config DB schema
@@ -34,10 +72,61 @@ latency              = 1*10DIGIT  ; server network latency in MS, -1 for connect
 ; Key
 config_key           = 'config'  ;  The configuration key
 ; Attributes
+enable                   = BOOLEAN   ; Enable Monitor feature
 time_window              = 1*5DIGIT  ; Monitor time window in minute, default is 5
 high_latency_threshold   = 1*5DIGIT  ; High latency threshold in ms, default is 20
 ```
 
+#### TACPLUS_MONITOR Table YANG model
+```yang
+module sonic-tacplus-monitor {
+    namespace "http://github.com/sonic-net/sonic-tacplus-monitor";
+    prefix tacplus-monitor;
+    yang-version 1.1;
+
+    organization
+        "SONiC";
+
+    contact
+        "SONiC";
+
+    description
+        "SONiC TACACS server monitor";
+
+    revision 2023-10-18 {
+        description
+            "Initial revision.";
+    }
+
+    container sonic-tacplus-monitor {
+
+        container TACPLUS_MONITOR {
+            description "TACACS server monitor config table";
+
+            container Config {
+                leaf enable {
+                    description "Enable Monitor feature";
+                    type boolean;
+                    default true;
+                }
+
+                leaf time_window {
+                    description "Monitor time window in minute";
+                    type uint16;
+                    default 5;
+                }
+
+                leaf high_latency_threshold {
+                    description "High latency threshold in ms";
+                    type uint16;
+                    default 20;
+                }
+            }
+        }
+    }
+}
+```
+
 # 3 Limitation
 
 - Service priority change will have 1 minutes delay, this is because monit service will run profile every 1 minutes.
@@ -71,13 +160,16 @@ high_latency_threshold   = 1*5DIGIT  ; High latency threshold in ms, default is
 - TACACS+ monitor also will write warning message to syslog when following event happen:
     - Any server latency is -1, which means the server is unreachable.
     - Any server latency is bigger than high_latency_threshold.
-- Hostcfgd will monitor TACPLUS_SERVER_LATENCY table, and will re-generate TACACS config file when following event happen:
-    - Any server latency is -1, which means the server is unreachable.
-    - Any server latency is bigger than high_latency_threshold.
+- Hostcfgd will monitor TACPLUS_SERVER_LATENCY table and re-generate TACACS config file:
+    - Hostcfgd will cache last time table data, and compare with latest table data.
+    - Based on compare result, if found following change, hostcfgd will re-generate TACACS config file:
+        - Any server latency is -1, which means the server is unreachable.
+        - Any unreachable server become reachable.
+        - Any server latency is bigger than high_latency_threshold.
 - When hostcfgd generate TACACS config file, server priority calculated according to following rules:
     - Get server priority info from CONFIG_DB TACPLUS_SERVER table.
     - Change high latency server to 1, this is because 1 is the smallest priority, and SONiC device will use high priority server first.
-    - Un-reachable server will not include in TACACS config file.
+    - Un-reachable server will not include in TACACS config file. When server become reachable, hostcfgd will add server back.
     - If other server also has priority 1 in CONFIG_DB, change priority to 2
     - If other server priority is no 1, using original priority in CONFIG_DB