Troubleshooting
From CEDPS
Contents |
Goals of the CEDPS Troubleshooting Area
Provide tools and expertise for end-to-end troubleshooting of DOE science, including:
- applications
- execution services
- data movement services
- networks
- host systems
Contact info Troubleshooting has its own mailing list; ceds-ts@mcs.anl.gov. Click here to subscribe. But most traffic is on the main CEDPS mailing list: ceds@mcs.anl.gov
Documents
- Grid Logging Best Practices: Recommendations for generating "good" log files, i.e. that are helpful for troubleshooting.
- Latest Best Practices Guide
- Latest cheat-sheet (condensed summary of the document)
- Specific recommendations for WS-RF Log Messages
- Accepted SC08 poster abstract (PDF)(MS-Word)
Software
Development funded by CEDPS:
Log collection:
- Survey of log collection tools
- syslog-ng, Open Source Edition
- Snare
- Rsyslog
Collaborations
Globus Toolkit
GT4.0 and earlier:
- Existing Log Examples with suggested modifications
- Sample GridFTP log levels
- Current WS Log example
- Simple python script to generate log entries
- log spreadsheets. Excel spreadsheet(s) summarizing log messages you see for some selected high-level operations. Easier to read than the raw logs.
GridFTP
Pegasus
Pegasus is scientific workflow management middleware from a group at USC/ISI.
- Pegasus/Kickstart Log Processing
- Pegasus Sample Queries
- Comparison of Kickstart and DAGMan logs
- Calculating DAGMan delay
- Notes from discussions with the Pegasus team
- NetLogger page on Pegasus wiki
PDSF
OSG
STAR/Tech-X
- Techx portal integration
- TechX meeting notes
- SRM from PDSF to BNL for STAR (7/31/2008)
- PDSF - BNL bandwidth measurements
