Overcoming Scalability Challenges for Tool Daemon Launching

Conference · Fri Feb 15 00:00:00 EST 2008

DOI:https://doi.org/10.1109/ICPP.2008.63· OSTI ID:944365

Ahn, D H; Arnold, D C; de Supinski, B R; Lee, G L; Miller, B P; Schulz, M

Many tools that target parallel and distributed environments must co-locate a set of daemons with the distributed processes of the target application. However, efficient and portable deployment of these daemons on large scale systems is an unsolved problem. We overcome this gap with LaunchMON, a scalable, robust, portable, secure, and general purpose infrastructure for launching tool daemons. Its API allows tool builders to identify all processes of a target job, launch daemons on the relevant nodes and control daemon interaction. Our results show that Launch-MON scales to very large daemon counts and substantially enhances performance over existing ad hoc mechanisms.

• Research Organization: Lawrence Livermore National Lab. (LLNL), Livermore, CA (United States)
• Sponsoring Organization: USDOE
• DOE Contract Number: W-7405-ENG-48
• OSTI ID: 944365
• Report Number(s): LLNL-CONF-401480; TRN: US200902%%726
• Resource Relation: Conference: Presented at: INTERNATIONAL CONFERENCE ON PARALLEL PROCESSING, Portland, OR, United States, Sep 08 - Sep 12, 2008
• Country of Publication: United States
• Language: English

Related Subjects

99 GENERAL AND MISCELLANEOUS
BUILDERS
LAUNCHING
PARALLEL PROCESSING
PERFORMANCE
TARGETS