Skip site navigation (1)Skip section navigation (2)
Date:      Sat, 1 Apr 2017 16:41:06 +0300
From:      Alexey Pereklad <technical@at-hacker.in>
To:        freebsd-questions@freebsd.org
Subject:   Re: Can't replace disk on LSI HBA
Message-ID:  <9682cbdf-6d9a-59c1-e836-0967a094534c@at-hacker.in>
In-Reply-To: <20170401151424.56b44a2c.freebsd@edvax.de>
References:  <013e2507-dbc5-8d75-d68d-ac139ed0b0bb@at-hacker.in> <20170401151424.56b44a2c.freebsd@edvax.de>

next in thread | previous in thread | raw e-mail | index | archive | help
Hi.

The same thing -- after "camcontrol stop da16" all the disks detached :-(


01.04.2017 16:14, Polytropon пишет:
> On Sat, 1 Apr 2017 15:42:09 +0300, Alexey Pereklad wrote:
>> I have FreeBSD 11.0 (11.0-RELEASE-p1 #1 r312693) here. Trying to test
>> hot disk replacement on LSI 9341-8i
>>
>> Put disk into offline state does not work:
>>
>>    # MegaCli -pdoffline -physdrv\[29:23\] -a1 -NoLog
>> Adapter: 1: Failed to change PD state at EnclId-29 SlotId-23.
>> Exit Code: 0x01
>>
>> I've tried to remove da9 device without putting it offline. Here is the
>> log:
>>
>> # grep detach /var/log/messages
>>
>> Apr  1 14:31:02 dc1-nfs-06 kernel: da9: <ATA WDC WD2004FBYZ-0 RR04> s/n
>> WD-WMC6N0D3S29A detached
>> Apr  1 14:31:03 dc1-nfs-06 kernel: ses0: <LSI SAS3x28 0401> detached
>> Apr  1 14:31:03 dc1-nfs-06 kernel: da2: <ATA WDC WD2000F9YZ-0 1A01> s/n
>> WD-WMC5C0D72CZC detached
>> Apr  1 14:31:03 dc1-nfs-06 kernel: da4: <ATA WDC WD2000F9YZ-0 1A01> s/n
>> WD-WMC5C0D18DVW detached
>> Apr  1 14:31:03 dc1-nfs-06 kernel: da6: <ATA WDC WD2000F9YZ-0 1A01> s/n
>> WD-WMC5C0D366LH detached
>> Apr  1 14:31:03 dc1-nfs-06 kernel: ses1: <LSI SAS3x40 0601> detached
>> Apr  1 14:31:03 dc1-nfs-06 kernel: da11: <ATA WDC WD2000F9YZ-0 1A02> s/n
>> WD-WMC1P0F4DY9M detached
>> Apr  1 14:31:03 dc1-nfs-06 kernel: ses0: <LSI SAS3x28 0401> detached
>> Apr  1 14:31:03 dc1-nfs-06 kernel: da3: <ATA WDC WD2000F9YZ-0 1A01> s/n
>> WD-WMC5C0D20R1C detached
>> Apr  1 14:31:03 dc1-nfs-06 kernel: s/n WD-WCC1P5ZVVPTD0): detached
>> Apr  1 14:31:03 dc1-nfs-06 kernel: da7: <ATA WDC WD2004FBYZ-0 RR04> s/n
>> WD-WMC6N0D11C71 detached
>> Apr  1 14:31:03 dc1-nfs-06 kernel: da15: <ATA WDC WD2004FBYZ-0 RR04> s/n
>> WD-WMC6N0D9F2KA detached
>> Apr  1 14:31:03 dc1-nfs-06 kernel: da8: <ATA WDC WD2004FBYZ-0 RR04> s/n
>> WD-WMC6N0D3PJDS detached
>> Apr  1 14:31:03 dc1-nfs-06 kernel: da10: <ATA WDC WD2004FBYZ-0 RR04> s/n
>> WD-WMC6N0E30R5D detached
>> Apr  1 14:31:03 dc1-nfs-06 kernel: da12: <ATA WDC WD2004FBYZ-0 RR04> s/n
>> WD-WMC6N0D97TPZ detached
>> Apr  1 14:31:58 dc1-nfs-06 kernel: ses0: <LSI SAS3x28 0401> detached
>> Apr  1 14:31:58 dc1-nfs-06 kernel: da4: <ATA WDC WD2000F9YZ-0 1A01> s/n
>> WD-WMC5C0D18DVW detached
>> Apr  1 14:31:58 dc1-nfs-06 kernel: da2: <ATA WDC WD2000F9YZ-0 1A01> s/n
>> WD-WMC5C0D72CZC detached
>> Apr  1 14:31:58 dc1-nfs-06 kernel: da6: <ATA WDC WD2000F9YZ-0 1A01> s/n
>> WD-WMC5C0D366LH detached
>> Apr  1 14:31:58 dc1-nfs-06 kernel: da8: <ATA WDC WD2004FBYZ-0 RR04> s/n
>> WD-WMC6N0D3PJDS detached
>> Apr  1 14:31:58 dc1-nfs-06 kernel: ses1: <LSI SAS3x40 0601> detached
>> Apr  1 14:31:58 dc1-nfs-06 kernel: da5: <ATA WDC WD2000F9YZ-0 1A02> s/n
>> WD-WCC1P5ZVVPTD detached
>> Apr  1 14:31:58 dc1-nfs-06 kernel: da3: <ATA WDC WD2000F9YZ-0 1A01> s/n
>> WD-WMC5C0D20R1C detached
>> Apr  1 14:31:58 dc1-nfs-06 kernel: da9: <ATA WDC WD2000F9YZ-0 1A02> s/n
>> WD-WMC1P0F4DY9M detached
>> Apr  1 14:31:58 dc1-nfs-06 kernel: da7: <ATA WDC WD2004FBYZ-0 RR04> s/n
>> WD-WMC6N0D11C71 detached
>> Apr  1 14:31:58 dc1-nfs-06 kernel: da10: <ATA WDC WD2004FBYZ-0 RR04> s/n
>> WD-WMC6N0E30R5D detached
>> Apr  1 14:31:58 dc1-nfs-06 kernel: da12: <ATA WDC WD2004FBYZ-0 RR04> s/n
>> WD-WMC6N0D97TPZ detached
>> Apr  1 14:31:58 dc1-nfs-06 kernel: da11: <ATA WDC WD2004FBYZ-0 RR04> s/n
>> WD-WMC6N0D9F2KA detached
>>
>> As you can see, all the drives has been detached. So is there any way to
>> replase failed drives on FreeBSD with LSI HBAs without reboot?
> Did you try to use "camcontrol"? As the devices are represented
> to regular "da" devices to the system by the HBA, maybe you could
> use "camcontrol stop" to detach the disk from the system prior
> to physically removing it? Check "camcontrol devlist" for the
> correct LUN of the disk you wish to replace.
>
> I haven't tested this approach, though. See "man camcontrol" for
> details and examples.
>
>




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?9682cbdf-6d9a-59c1-e836-0967a094534c>