[202012][fast-reboot] Remove FLEX_COUNTER_TABLE from config_db.json b…

…efore reboot (#1774)

Remove FLEX_COUNTER_TABLE from config_db.json before fast-reboot to allow delaying FLEX counter polling after fast-reboot.
Delaying FLEX counter polling is important to keep fastboot dataplane downtime under 30s.
In the going down path, add a step to modify config_db.json - remove the key:value for FLEX_COUNTER_TABLE table.
Repro'd the issue in the latest 202012 image.
With the fix, the counter polling is delayed and downtime is back to normal.

scripts/fast-reboot

@@ -20,6 +20,7 @@ PLATFORM=$(sonic-cfggen -H -v DEVICE_METADATA.localhost.platform)
 PLATFORM_PLUGIN="${REBOOT_TYPE}_plugin"
 LOG_SSD_HEALTH="/usr/local/bin/log_ssd_health"
 SSD_FW_UPDATE="ssd-fw-upgrade"
+CONFIG_DB_FILE="/etc/sonic/config_db.json"
 TAG_LATEST=yes
 
 # Require 100M available on the hard drive for warm reboot temp files,
@@ -512,7 +513,6 @@ if [[ "$REBOOT_TYPE" = "fast-reboot" ]]; then
     # Dump the ARP and FDB tables to files also as default routes for both IPv4 and IPv6
     # into /host/fast-reboot
     DUMP_DIR=/host/fast-reboot
-    CONFIG_DB_FILE=/etc/sonic/config_db.json
     mkdir -p $DUMP_DIR
     FAST_REBOOT_DUMP_RC=0
     /usr/local/bin/fast-reboot-dump.py -t $DUMP_DIR || FAST_REBOOT_DUMP_RC=$?
@@ -524,7 +524,7 @@ if [[ "$REBOOT_TYPE" = "fast-reboot" ]]; then
 
     FILTER_FDB_ENTRIES_RC=0
     # Filter FDB entries using MAC addresses from ARP table
-    /usr/local/bin/filter_fdb_entries -f $DUMP_DIR/fdb.json -a $DUMP_DIR/arp.json -c $CONFIG_DB_FILE || FILTER_FDB_ENTRIES_RC=$?
+    /usr/local/bin/filter_fdb_entries -f $DUMP_DIR/fdb.json -a $DUMP_DIR/arp.json -c ${CONFIG_DB_FILE} || FILTER_FDB_ENTRIES_RC=$?
     if [[ FILTER_FDB_ENTRIES_RC -ne 0 ]]; then
         error "Failed to filter FDb entries. Exit code: $FILTER_FDB_ENTRIES_RC"
         unload_kernel
@@ -679,6 +679,14 @@ then
   systemctl stop "$service_name"
 fi
 
+if [[ "$REBOOT_TYPE" = "fast-reboot" ]]; then
+    # Remove FLEX_COUNTER_TABLE from config_db.json
+    # This is done so that in fast-reboot recovery path, FLEX_COUNTER polling is delayed.
+    # Delayed FLEX_COUNTER polling is an attempt keep dataplane downtime below 30s threshold
+    jq --indent 4  'del(.FLEX_COUNTER_TABLE)' ${CONFIG_DB_FILE} > ${CONFIG_DB_FILE}.new
+    mv ${CONFIG_DB_FILE}.new ${CONFIG_DB_FILE}
+fi
+
 # Update the reboot cause file to reflect that user issued this script
 # Upon next boot, the contents of this file will be used to determine the
 # cause of the previous reboot