1 / 18

Grid Monitoring

Grid Monitoring. Warren Matthews and Adil Hasan (SLAC) Internet2 HENP BOF / ESCC ANL October 2001. Overview. SLAC IEPM since 1995 WAN Backbones typically perform very well Monitoring for PPDG Requires finer grained monitoring End-to-end is required

quincy
Download Presentation

Grid Monitoring

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. Grid Monitoring Warren Matthews and Adil Hasan (SLAC) Internet2 HENP BOF / ESCC ANL October 2001.

  2. Overview • SLAC IEPM since 1995 • WAN Backbones typically perform very well • Monitoring for PPDG • Requires finer grained monitoring • End-to-end is required • Provide statistics on network performance • Conduct analysis on trends for resource allocation • Contribute real data to a PPDG Monitoring system

  3. Global Grid Forum (GGF) • Several relevant WGs • Performance Working Group/Area • Network Working Group • Grid Information Services • Grid Monitoring Architecture (GMA) • GWD-Perf-16-1 • GMA implementation at SLAC

  4. GMA Terminology Consumer Event Publication Information Directory Service Events Producer

  5. Directory Service - LDAP

  6. Producers o=grid-mon dc=domain component dc=ppdg dc=SDSC dc=UWisc dc=Caltech dc=LBNL dc=BNL dc=JLAB dc=SLAC dc=ANL dc=FNAL hn=Oceanus.SLAC.Stanford.EDU hn=Jlab7.Jlab.ORG PingER

  7. Select remote site Select Start Time Select End Time

  8. Production Data • PingER data has been put into LDAP • required platform for making resource decisions • LDAP optimized for reading • Too dynamic, overwhelm server • FNAL (archive site) implementing MySQL • Not near real-time • Historic data required for alarm system • Common interface is useful

  9. Network Monitors

  10. GIMI • Global Internet Measurement Infrastructure • Platform for many measurements • Not tied to one-groups set of measurements • Similar feel to GMA • Possibly access to otherwise restricted data (eg SNMP) due to stringent access-control

  11. Consumers BNL JLAB FNAL SLAC LBNL Central Directory Service ANL UWisc Caltech SDSC Web Page Trouble-Shooting Tool Resource Broker

  12. Median Average RTT = 71.00ms Inter Quartile Range = 2.00ms

  13. Tools • Traceroute • Router details • Pathchar • Bottleneck bandwidth • Iperf (Enable) • Chirp (INCITE) • characterize network service

  14. Other Work • Euro-Grid • Globus/MDS • Standards • IETF/IPPM • Performance metrics • New implementations • GGF/Network WG

  15. Further Work • Production Monitoring • Directory Service • More Data • Analysis Tools • Integrate with Resource Brokering

  16. Future Work • Network Weather Service • Servers and/or integrate into GIMI • Host monitoring should be registered • Unknown timescale • Unfunded • Looking for volunteer effort

  17. Conclusions • Finer-grained monitoring is essential • Required service for the Grids • Can the I2 HENP group help ?

  18. Any Questions ?

More Related