Created attachment 463670 [details] patch Description of problem: with certain Atom based mother boards USB drive based installs fail to pick kickstart file from USB DVD drives the first time, There is a need to retry 2 to 3 times. Version-Release number of selected component (if applicable): EL 5.5 How reproducible: frequent Actual results: Before applying patch: 19:32:03 INFO : getting kickstart file 19:32:03 INFO : getting kickstart file from first CDROM <-- fails here 19:32:16 INFO : getting kickstart file from first CDROM 19:32:16 INFO : getFileFromBlockDevice(scd0, /ks.cfg) 19:32:23 INFO : Searching for file on path /tmp/mnt//ks.cfg 19:32:23 INFO : file copied to /tmp/ks.cfg 19:32:23 INFO : setting up kickstart 19:32:23 INFO : kickstartFromCD Expected results: 19:43:59 INFO : getting kickstart file 19:43:59 INFO : getting kickstart file from first CDROM 19:43:59 DEBUG : sleeping to wait for a USB CD-ROM 19:44:01 DEBUG : sleeping to wait for a USB CD-ROM 19:44:03 INFO : getFileFromBlockDevice(scd0, /ks.cfg) Additional info: related bug - Bug 460566 - kickstart file on USB drive not found on first attempt Patch from customer attached ( Sandeep K Shandilya from cisco)
The one comment I have on the attached patch is that I would prefer to probe before sleeping in the loop. There's no reason to penalize everyone with a sleep when it's only going to be the USB people who require it.
From the Logs I am able to see that in "bad" logs the usb-storage module is being loaded after we start looking for kickstart file. logs sony-good 18:48:11 INFO : inserted /tmp/usb-storage.ko 18:48:11 INFO : removed usb-storage from modprobe.conf 18:48:11 INFO : load module set done 18:48:18 INFO : load module set done 18:48:18 INFO : getting kickstart file 18:48:18 INFO : getting kickstart file from first CDROM 18:48:18 DEBUG : sleeping to wait for a USB CD-ROM 18:48:20 INFO : getFileFromBlockDevice(scd0, /ks.cfg) 18:48:25 INFO : Searching for file on path /tmp/mnt//ks.cfg 18:48:26 INFO : file copied to /tmp/ks.cfg 18:48:26 INFO : setting up kickstart 18:48:26 INFO : kickstartFromCD 18:48:26 DEBUG : going to set language to en_US.UTF-8 18:48:26 INFO : setting language to en_US.UTF-8 18:48:26 INFO : starting to STEP_URL 18:48:26 INFO : trying to mount CD device scd0 sony-bad 18:56:42 INFO : load module set done 18:56:42 INFO : getting kickstart file 18:56:42 INFO : getting kickstart file from first CDROM 18:56:42 DEBUG : sleeping to wait for a USB CD-ROM 18:56:44 DEBUG : sleeping to wait for a USB CD-ROM ... 18:59:10 DEBUG : sleeping to wait for a USB CD-ROM 18:59:12 ERROR : No CDROM devices found! 18:59:14 ERROR : got to setupCdrom without a CD device 18:59:15 DEBUG : going to set language to en_US.UTF-8 18:59:15 INFO : setting language to en_US.UTF-8 18:59:15 INFO : 53 keymaps are available 18:59:32 INFO : modules to insert usb-storage 18:59:32 DEBUG : getModuleLocation: 2.6.18-194.el5/i686 18:59:32 INFO : loaded usb-storage from /modules/modules.cgz 18:59:38 DEBUG : slept 6 seconds 18:59:39 INFO : inserted /tmp/usb-storage.ko 18:59:39 INFO : load module set done 18:59:39 INFO : starting to STEP_URL 18:59:39 INFO : trying to mount CD device scd0 From the Logs I am able to see that in "bad" logs the usb-storage module is being loaded pretty close to the point where we start looking for kickstart file. samsung-good 19:17:18 INFO : reloading module usb-storage 19:17:18 INFO : modules to insert usb-storage 19:17:18 DEBUG : getModuleLocation: 2.6.18-194.el5/i686 19:17:18 INFO : loaded usb-storage from /modules/modules.cgz 19:17:19 INFO : inserted /tmp/usb-storage.ko 19:17:19 INFO : removed usb-storage from modprobe.conf 19:17:19 INFO : load module set done 19:17:26 INFO : load module set done 19:17:26 INFO : getting kickstart file 19:17:26 INFO : getting kickstart file from first CDROM 19:17:26 DEBUG : sleeping to wait for a USB CD-ROM 19:17:28 DEBUG : sleeping to wait for a USB CD-ROM 19:17:30 INFO : getFileFromBlockDevice(scd0, /ks.cfg) 19:17:33 INFO : Searching for file on path /tmp/mnt//ks.cfg 19:17:33 INFO : file copied to /tmp/ks.cfg 19:17:33 INFO : setting up kickstart 19:17:33 INFO : kickstartFromCD 19:17:33 DEBUG : going to set language to en_US.UTF-8 19:17:33 INFO : setting language to en_US.UTF-8 19:17:33 INFO : starting to STEP_URL 19:17:33 INFO : trying to mount CD device scd0 samsung-bad 19:09:48 INFO : unloading module usb-storage 19:09:48 INFO : going to rmmod usb-storage 19:09:48 INFO : inserted /tmp/e1000e.ko 19:09:49 INFO : inserted /tmp/libata.ko 19:09:52 INFO : inserted /tmp/ahci.ko 19:09:53 INFO : inserted /tmp/ata_piix.ko 19:09:53 INFO : reloading module usb-storage 19:09:53 INFO : modules to insert usb-storage 19:09:53 DEBUG : getModuleLocation: 2.6.18-194.el5/i686 19:09:53 INFO : loaded usb-storage from /modules/modules.cgz 19:09:54 INFO : inserted /tmp/usb-storage.ko 19:09:54 INFO : removed usb-storage from modprobe.conf 19:09:54 INFO : load module set done 19:10:01 INFO : load module set done 19:10:01 INFO : getting kickstart file 19:10:01 INFO : getting kickstart file from first CDROM 19:10:26 INFO : getting kickstart file from first CDROM 19:10:26 INFO : getFileFromBlockDevice(scd0, /ks.cfg) 19:10:32 INFO : Searching for file on path /tmp/mnt//ks.cfg 19:10:32 INFO : file copied to /tmp/ks.cfg 19:10:32 INFO : setting up kickstart 19:10:32 INFO : kickstartFromCD 19:10:32 DEBUG : going to set language to en_US.UTF-8 19:10:32 INFO : setting language to en_US.UTF-8 19:10:32 INFO : starting to STEP_URL 19:10:32 INFO : trying to mount CD device scd0 A race condition with module loading ?
This request was evaluated by Red Hat Product Management for inclusion in a Red Hat Enterprise Linux maintenance release. Product Management has requested further review of this request by Red Hat Engineering, for potential inclusion in a Red Hat Enterprise Linux Update release for currently deployed products. This request is not yet committed for inclusion in an Update release.
Hello, can you please attach the complete anaconda.log files from times when the device was and was not found on the first attempt (from a non-patched loader)? Thanks.
Thank you for the patch, it is waiting for a review now: https://www.redhat.com/archives/anaconda-devel-list/2011-March/msg00057.html
Fixed by e51768a50d92558b285e0e37de562daec9538c91.
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 therefore 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-2011-0984.html