Login
[x]
Log in using an account from:
Fedora Account System
Red Hat Associate
Red Hat Customer
Or login using a Red Hat Bugzilla account
Forgot Password
Login:
Hide Forgot
Create an Account
Red Hat Bugzilla – Attachment 158556 Details for
Bug 239656
GFS2: gfs2_tool layout not implemented: regression from gfs1
[?]
New
Simple Search
Advanced Search
My Links
Browse
Requests
Reports
Current State
Search
Tabular reports
Graphical reports
Duplicates
Other Reports
User Changes
Plotly Reports
Bug Status
Bug Severity
Non-Defaults
|
Product Dashboard
Help
Page Help!
Bug Writing Guidelines
What's new
Browser Support Policy
5.0.4.rh83 Release notes
FAQ
Guides index
User guide
Web Services
Contact
Legal
This site requires JavaScript to be enabled to function correctly, please enable it.
Project proposal
proposal.txt (text/plain), 2.44 KB, created by
Andy Price
on 2007-07-04 21:31:30 UTC
(
hide
)
Description:
Project proposal
Filename:
MIME Type:
Creator:
Andy Price
Created:
2007-07-04 21:31:30 UTC
Size:
2.44 KB
patch
obsolete
>Project Description > >GFS2 is the new version of the Global File System which is a high performance cluster file system. It is being developed by Red Hat and has been included in the main line Linux kernel since version 2.6.16. The lead Red Hat developer of GFS2 is Steven Whitehouse who is based in the Technium building on the Swansea University campus. > >The idea of this project is to extend existing block layer file system debugging tools with information gathered from the GFS2 libraries, or otherwise, in order to lay the foundations for creating performance analysis tools for a GFS2 file system which is shared between many nodes. In order to do this it will be necessary to become familiar with how the cluster file system works, find out what information is needed in order to analyse the file system's performance from what is available and to write or extend the necessary libraries and tools that will allow the performance analysis tools to be created. There is also scope to generalise the tools to enable them to provide analysis for other block based file systems. > >The envisaged end product would be a tool that could be used by, say, a company employee to visualise the performance of their GFS2 cluster. At first glance it seems that there is a lot of groundwork to lay before an analysis tool could be implemented so it may take the form of proof-of-concept or a basic test program for the tools developed as the main part of the project. Other possibilities include writing a python interface for the GFS2 library, libgfs2, to make it easier for others to develop performance tools with it, implementing file system "snapshotting", providing the gathered data in different formats for existing visualisation tools etc. > >Project Features > >For this project I will be working in close contact with Red Hat and contributing to a worthwhile project that is under heavy development. I am already involved in open source software in my spare time so I have a keen interest in working on a project of this nature. The target market would initially be developers and users of GFS2 but it could extend further, depending on how much is achieved and which direction it takes. > >From initial research, it seems that there are some challenging problems to be dealt with in parsing the appropriate data from the file system and dealing with merged data in the existing block layer debugging information (from the elevator algorithm) so there is plenty of scope to make this a good project.
You cannot view the attachment while viewing its details because your browser does not support IFRAMEs.
View the attachment on a separate page
.
View Attachment As Raw
Actions:
View
Attachments on
bug 239656
: 158556