I’ve already identified b as well and it’s completely orthogonal to the bug in question. When you install the new card and create the new virtual disks, the controller will present a popup screen suggesting that you initialize the disks, we had existing data on our servers to I choose NOT to do that, and we have not seen any negative side effects of not doing so. The latest mainline kernel is v3. This is a regression since those server worked in precise and work with precise old kernel. I proposed the final patch but I’m unable to prove that SIGKILL sent by systemd-udevd’s 30 seconds timeout is the trigger of this problem, for I don’t have a real machine which takes very long time upon initialization. I still didn’t tested the firmware upgrade.
|Date Added:||21 February 2007|
|File Size:||67.92 Mb|
|Operating Systems:||Windows NT/2000/XP/2003/2003/7/8/10 MacOS 10/X|
|Price:||Free* [*Free Regsitration Required]|
I didn’t tested it to keep a machine which exhibit the bug Debian kernel 10688e rejected IO to offline devices in case array becomes downgraded ie one disk failed.
If not, is the source for some command line tool available which can report this status? Not aware of any method of detecting the status for this RAID. Now we proved that systemd-udevd’s 30 seconds timeout is the trigger of this problem.
We implemented the rootdelay workaround and that let the system boot without problems. No idea about athough, your comment is interesting. Probably the most interesting one, dmesg was complete include very first messages.
This was not the case on at least one Dell PowerEdge machine. Fixed not processing task management IOCB response status git bisect good bbb83f11b3 58ac94eddd cdc51ea good: Tim Gardner timg-tpi on It is NOT working, it has the issue. Because of timezone difference we can only test one kernel per day, to speed up the bisect, I’ve done one by myself, the result is the following: When the utility finishes creating the volume, the main screen reappears.
September 29th, 3. The kernel generated too much logs for IPMI serial console which generated too much garbageso I switched to a real serial console and at kbauds. Console output with kthread-defer-leaving v2 patch applied on ee Edit On Debian based system, simply run:. Same source requirements as the CLI tools.
Can you test the following kernels and report back? Well we replaced our LSI cards with the Perc ones without any issues.
In order to bisect, we need to narrow down further and identify which v3. From my bisect, the commit which introduced the error is 8ceafbfa91ffbdb b2afaea5c24ccb5 19ffb8b Pressing control-D resume normal boot So the issue is that with the mandatory rootdelay greater that 30 default value I thinkthe disk are not detected due to the error kinux above.
Well if you provide me with a little bit more detail I might be able to offer up some help…however as the article states, turning in the write cache option via the admin tools really has little to no actual impact on the performance that you will receive while using this card. I read this thread. This change was made by a bot.
I will propose this patch for 3. Select Save changes then exit this menu, and then press Enter. Join Date Sep Beans 7. During my bisect, I always booted 3 times on good kernel to make sure it was not by “luck” that the kernel worked. Find More Posts by Tinkster. lwi
Thanks all for your work.