refactor(monitoring): simplify notification system and remove auto-cleanup

- Replace state-based notifications with direct alert functions
- Remove auto-cleanup functionality from disk monitoring and configuration
- Simplify lock acquisition/release across all monitoring scripts
- Add execute_hana_sql helper functions for consistent SQL execution
- Remove state file tracking in favor of direct file operations
- Standardize error handling with exit codes on critical failures
- Clean up hana.conf by removing unused auto-delete directory settings
This commit is contained in:
2026-03-12 21:52:49 +01:00
parent 5a92bc4e93
commit cf5b81889d
8 changed files with 183 additions and 391 deletions

View File

@@ -1,10 +1,8 @@
#!/bin/bash
#
# =============================================================================
# sld_watchdog.sh - Monitors SLD service health and restarts if needed
# Optimized for better error handling and reliability
#
# =============================================================================
# Get script directory and name
SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
SCRIPT_NAME="sld_watchdog"
@@ -15,14 +13,12 @@ source "$SCRIPT_DIR/hana_lib.sh"
# SLD-specific configuration
SLD_URL="${SLD_URL:-https://localhost:40000/sld/sld0100.svc}"
SLD_TIMEOUT="${SLD_TIMEOUT:-5}"
SLD_LOCK_FILE="/tmp/hana_sld_watchdog.lock"
# Acquire lock using library function
LOCK_FILE=$(acquire_lock "$SCRIPT_NAME")
if [ $? -ne 0 ]; then
# Acquire lock
if ! acquire_lock "$SCRIPT_NAME"; then
exit 0
fi
trap 'release_lock "$LOCK_FILE"' EXIT
trap 'release_lock "$SCRIPT_NAME"' EXIT
# Function to check SLD health
# Returns HTTP status code or "0" for connection errors
@@ -71,7 +67,6 @@ main() {
# 200 OK or 401 Unauthorized indicate the Tomcat layer is actively processing
if [[ $http_status == 200 || $http_status == 401 ]]; then
log_message "$SCRIPT_NAME" "[OK] SLD service healthy (HTTP Status: $http_status)"
send_notification_if_changed "$SCRIPT_NAME" "sld_status" "SLD Service" "SLD service is healthy (HTTP: $http_status)" "false" "OK"
return 0
fi
@@ -86,14 +81,12 @@ main() {
log_message "$SCRIPT_NAME" "[ALERT] SLD dead or hung (${status_detail}). Restarting service..."
# Send notification
send_notification_if_changed "$SCRIPT_NAME" "sld_down" "SLD Service Critical" \
"SLD service is down (${status_detail}). Restarting ${SLD_URL}" "true" "SLD_DOWN"
send_alert "$SCRIPT_NAME" "SLD Service Critical" "SLD service is down (${status_detail}). Restarting ${SLD_URL}"
# Restart the service
if ! restart_sld_service; then
log_message "$SCRIPT_NAME" "[CRITICAL] Failed to restart SLD service"
send_notification_if_changed "$SCRIPT_NAME" "sld_restart_failed" "SLD Service Critical" \
"Failed to restart SLD service" "true" "RESTART_FAILED"
send_alert "$SCRIPT_NAME" "SLD Service Critical" "Failed to restart SLD service"
return 1
fi
@@ -106,10 +99,6 @@ main() {
if [[ $recovery_status == 200 || $recovery_status == 401 ]]; then
log_message "$SCRIPT_NAME" "[RECOVERY] SLD service recovered successfully (HTTP Status: $recovery_status)"
send_notification_if_changed "$SCRIPT_NAME" "sld_down" "SLD Service" \
"SLD service recovered (HTTP: $recovery_status)" "false" "OK"
send_notification_if_changed "$SCRIPT_NAME" "sld_restart_failed" "SLD Service" \
"Service recovered successfully" "false" "OK"
else
local recovery_detail
if [ "$recovery_status" == "0" ]; then
@@ -118,8 +107,7 @@ main() {
recovery_detail="HTTP Status: $recovery_status"
fi
log_message "$SCRIPT_NAME" "[CRITICAL] SLD service failed to recover after restart (${recovery_detail})"
send_notification_if_changed "$SCRIPT_NAME" "sld_down" "SLD Service Critical" \
"SLD service FAILED to recover after restart (${recovery_detail})" "true" "RECOVERY_FAILED"
send_alert "$SCRIPT_NAME" "SLD Service Critical" "SLD service FAILED to recover after restart (${recovery_detail})"
return 1
fi