Monitoring as a Service
- 1. Monitoring As A Service Vincent.Tacquet@synergics.be Quality & Architecture Manager Datacenter Manager www.synergics.be
- 3. Checks Active Checks Active checks are initiated by the Monitoring Daemon, the host to be checked returns the result (ping, memory usage, cpu load, tcp port, snmp query, ...). Passive Checks Passive checks are initiated by the host to be checked and the results are pushed to the Monitoring Daemon (snmp trap, scripts). www.synergics.be 3
- 5. Results Results are stored in RRD files RRD = round robin database (a fixed number of values – 5 years history stored) RRD files are imported into an Oracle Database for Reporting / Data Mining / Business Intelligence www.synergics.be 5
- 7. www.synergics.be 7
- 8. www.synergics.be 8
- 9. www.synergics.be 9
- 10. Drraw www.synergics.be 10
- 11. WeatherMap www.synergics.be 11
- 12. NagVis (location based) www.synergics.be 12
- 13. www.synergics.be 13
- 14. NagVis (function based) www.synergics.be 14
- 15. www.synergics.be 15
- 17. Integration www.synergics.be 17
- 18. www.synergics.be 18
- 20. Redundancy www.synergics.be 20
- 21. Reporting www.synergics.be 21
- 22. €xtra Services Configuration (standard and/or custom checks, graphs, visual design, network topo) Escalations Notifications Trending SLA’s Reporting / Data Mining / BI www.synergics.be 22
- 23. www.synergics.be 23
Editor's Notes
- #3: General Overview of all services available on our Monitoring As A Service product.
- #4: Custom checks can be madeReturn value (0: OK, 1: WARNING, 2: CRITICAL, 3: UNKNOWN)Passive checks (if no result was pushed for the last x minutes/hours : the monitoring daemon can put a service into CRITICAL)DEMO: cdimp0 (cpuload), nzwnups-01a0019 (traps)
- #5: Host left pane (mind the brand icon)Service right pane (mind the blue icon | attemts | status information)DEMO: cdimp0 (cpuload), nzwnups-01a0019 (traps)
- #6: RRD files are +- 5Mb (depending on number of save points and history)Oracle Database is currently 790 Gb large and has 7 billion records(slide 23)
- #7: Not all checks have the blue icon (not useful for certificate check)
- #8: Green is AverageRed is Max (on hourly basis) Black is Previous (Day/Week)
- #10: Trending, 4 possibilities, each with a 100% crossing calculation to a date + visualy (when will the DISK be full, when is the CPU saturated)DEMO: sdtcasoaccweb1 (diskusage-c)
- #11: Allows you to create your own graphs (example of Traffic with Users on Citrix farm; example of temperature) to analyze further.Allows you to analyze the impact of 10 extra users on CPU Load / Memory UsageDashboards are possible too (grouping of multiple graphs)
- #12: Network diagram with bandwidth usage (up/down), allows you to visualize bottlenecks
- #13: Location based representationClickable icons (shows live information of the host)
- #14: Datacenter floorplan
- #15: Function based representationShows flowsDiagrams can be made in anything you like (Visio, other flowchart tool, Photoshop, ...)
- #16: Example of a flowchart created for our customer Attentia Soc Sec
- #17: Notifications are possible in all kind of flavours (sms, mail, twitter)Same info on SMS or on Mail (on SMS short words)30 second delay send vs. receive (we call our provider by means of an analogue line / redundant)
- #18: Integration in our CMDB and Ticketing systemDEMO: nzwncsw-01t201v (iface0014octets)
- #20: Monthly report for our customer Attentia Soc Sec (SAE3402 compliancy) with the uptimes.
- #21: Second Nagios in DIA datacenterWhen First Nagios in ZWN datacenter is not responding, the Second Nagios will takeover and notify Monitoring CompetenceRRDB only in ZWN datacenter, storage replicated, not updated when in DR mode
- #22: Database RRDB (with 7 billion records), query via Access / Excel / TOAD or any other tool