Red Hat Bugzilla – Full Text Bug Listing
|Summary:||GFS2: Evaluate and implement missing gfs2_tool features|
|Product:||Red Hat Enterprise Linux 5||Reporter:||Robert Peterson <rpeterso>|
|Component:||gfs2-utils||Assignee:||Chris Feist <cfeist>|
|Status:||CLOSED ERRATA||QA Contact:||GFS Bugs <gfs-bugs>|
|Fixed In Version:||RHBA-2008-0350||Doc Type:||Bug Fix|
|Doc Text:||Story Points:||---|
|Last Closed:||2008-05-21 13:20:22 EDT||Type:||---|
|oVirt Team:||---||RHEL 7.3 requirements from Atomic Host:|
Description Robert Peterson 2007-10-25 13:35:25 EDT
Description of problem: This is an offshoot of bug #349601 comment #3: The gfs2_tool command has a bunch of commands that were commented out for RHEL5. That was done mostly because there were numerous ioctls pulled out of GFS2 in order to satisfy the upstream community. Most of these, however, have commands in gfs2_edit that do nearly the same thing. For example, gfs2_edit -p rindex can be said to replace gfs2_tool rindex and gfs2_edit -p jindex can be said to replace gfs2_tool jindex. However, I won't say that the gfs2_edit command prints the exact same format as the gfs2_tool. We should probably go through all those gfs2_tool commands, evaluate which ones are adequately covered by gfs2_edit and re-implement the ones that aren't. That is, if they can be done without re-implementing new ioctl code in gfs2 kernel code. Case in point: gfs2_tool df. Version-Release number of selected component (if applicable): RHEL5.1 How reproducible: Always Steps to Reproduce: gfs2_tool df Actual results: gfs2_tool: unknown action: df Expected results: Should work Additional info:
Comment 2 Robert Peterson 2007-11-08 15:10:08 EST
I've gone through the options that are missing from gfs2_tool and here is my analysis. The following is a list of "missing" options and where we stand: 1. Option: flush Description: Sync out dirty data for a file and drop its glock. Status: There's no good way to implement this today, except open and fsync. 2. Option: stat Description: Print gfs2_dinode for a file. Status: There's no good way to implement this today. The gfs2_edit command can do this, give the inode address. Right now you can't give it a path. 3. Option: getsb Description: Print out the gfs2 superblock Status: There is a command equivalent in gfs2_edit: gfs2_tool -p sb so I don't see the need to implement this. 4. Option: jindex Description: Print out the journal index Status: There is a command equivalent in gfs2_edit: gfs2_tool -p jindex so I don't see the need to implement this. 5. Option: rindex Description: Print out the resource group index file Status: There is a command equivalent in gfs2_edit: gfs2_tool -p rindex so I don't see the need to implement this. 6. Option: quota Description: Print out the quota file details Status: There is a command equivalent in gfs2_edit: gfs2_tool -p quota so I don't see the need to implement this. 7. Option: layout Description: Print out the layout of a file Status: There is a separate bugzilla record to implement this feature. 8. Option: df Description: Print out a bunch of information about the file system Status: Some customer might be using this. It might be nice to implement this. Conclusion: IMHO, the only missing option we should consider implementing at this time is the "df" command.
Comment 3 Robert Peterson 2007-11-13 17:22:11 EST
Created attachment 257481 [details] Patch to add the "df" functionality This is my initial attempt to add the "df" function back to gfs2_tool.
Comment 4 Robert Peterson 2007-11-14 12:16:04 EST
Created attachment 258441 [details] Better patch This version of the patch is the one I'm planning to commit. It has a few small improvements, such as the "df" help printing in the alphabetically correct section of the gfs2_tool -h output. I also eliminated some cruft from the previous patch.
Comment 5 Robert Peterson 2007-11-14 12:22:06 EST
The code was tested on system roth-01 including the man page, -h option, and new df options for both single and multiple file systems (both HEAD and RHEL5 versions). The code was committed to the HEAD and RHEL5 branches of CVS. I'm therefore changing the status to modified.
Comment 8 errata-xmlrpc 2008-05-21 13:20:22 EDT
An advisory has been issued which should help the problem described in this bug report. This report is therefore being closed with a resolution of ERRATA. For more information on the solution and/or where to find the updated files, please follow the link below. You may reopen this bug report if the solution does not work for you. http://rhn.redhat.com/errata/RHBA-2008-0350.html