Zabbix 선택 배경
- Considerations on IT Infrastructure Monitoring
HPE Experiences & Knowledge DB
-
Performance
성능
-
- CPU/Memory high utilization
- Network bandwidth usage
- Packet loss rate
- Interface error rate
- Number of tcp connections is anomaly high for this day of the week
- Aggregate throughput of core routers is low
Availability
가용성
-
- Free disk space is low
- System status is in warning/critical state
- Device temperature is too high / too low
- Power supply is in critical state
- Fan is in critical state
- No SNMP data collection
- Cluster status
Management
관리
-
- New components added or removed
- Network module is added, removed or replaced
- Firmware upgraded
- Device serial number has changed
- Interface changed to lower speed or half-duplex mode
- Configuration backup
- Monitoring Items for High Performance Computing
HPE Best Practice
-
- CPU
-
- Load average
- CPU idle/usage
- CPU utilization data per individual process
- Memory
-
- Free/used memory
- Swap/pagefile utilization
- Disk
-
- Space free/used
- Read and write I/O
- Service
-
- Process status/memory usage
- Service status (ssh, ldap, ftp, http)
- Windows service status
- DNS resolution
- TCP/UDP connectivity,
- TCP/UDP response time
- File
-
- File size/time, File exists
- Checksum
- RegExp search
- H/W
-
- Sensor reading
- BMC (HPE iLO, Dell iDRAC, etc)
- Temperature, Power (Watt)
- Chassis/Fan/Chipset/Drives/DIMM
- PCI/USB devices (Controller)
- Firmware/Driver (NIC, HBA, Controller)
- Other
-
- Log file monitoring
- Kernel (Max no. file, Max no. process)
- System uptime, Users connected
- Cluster file/process/service/package
- Zabbix
개요
- Zabbix
기능
| Data Collect | Collect From Any Source (OS level, LLD) Flexible Metric Collection (log file, event) Zabbix Agent (Push, Pull) Agent-Less Monitoring (SNMP, IPMI, SSH) Synthetic Monitoring (Web Application and APIs) Custom Collection Methods (CLI Util, External Scripts) Data Tansformation (JSON, XML, CSV) |
|---|---|
| Problem Detect | Smart Thresholds (Threshold) Trend Prediction (Forecast, Timeleft) Machine Learning (Insight) |
| Alert | Messaging Channels (E-mail, SMS, Communication Channel) Messages Customization (Customized Message of Issue) Escalation Scenarios (Notification -> Auto Action -> Run Remote Scripts -> Notification Auto-Remediation (Custom Action Scripts : Restart Services, Delete) |
| Visualization | Indivdual Dashboards Geo-Maps Infrastructure Maps Schedule Reports |
| Single Pane of Glass | Widget base multi page dashboards Multi Tenancy Inventory Information |
| Business Monitoring | Root Cause Analysis Business-Level Impact SLA Monitoring |
| Integration | Support out-of box templates for software and hardware vendors Support zabbix APIs |
| Security | Support TLS protocol Individual permissions User Authentication |
| Deployment | Easy Install Agent Network, Resource Discovery |
| Scalability | Umlimit Proxys High Availability |
- Zabbix
모니터링 대상
Network
| Network performance | Network health | Configuration changes |
|---|---|---|
|
|
|
| Template 지원 Vendor | ||
 |
||
Networks
-
Servers
- Zabbix 이용한 HPE ProLiant 모니터링 방안
Agent (by OS)
- Zabbix 이용한 HPE ProLiant Gen8/9 모니터링 방안
Agentless (by SNMP, IPMI) using Template
-
Agentless 모니터링 항목 예시 (SNMP & IPMI)
Excel 파일 “Zabbix_Template 전체 정리＂
-
HPE iLO SNMP OID(MIB) 값
Excel 파일 “MIB Public Gen10＂
-
HPE iLO IPMI Sensor ID 값
PDF 파일 “HPE iLO IPMI User Guide“
-
Zabbix Template 예시
Zabbix 홈페이지 내 각 vendor Template 게시판 https://share.zabbix.com/cat-server-hardware/hp
XML 파일 “Template HP iLO4 SNMP Agentless”, “Template HP DL380 Gen9 IPMI”
Excel 파일 “Zabbix Template HPE Server 분석”, “Zabbix Template Supermicro Server 분석”
- Zabbix 이용한 HPE ProLiant Gen8/9 모니터링 방안
HPE ProLiant Template 분석
- Zabbix 이용한 HPE ProLiant Gen10 모니터링 방안
IPMI 및 SNMP 통한 integration 방안
