From blikjeham at vierkleurenspreeuw.nl Wed Sep 24 10:46:23 2008 From: blikjeham at vierkleurenspreeuw.nl (blikjeham@vierkleurenspreeuw.nl) Date: Wed Sep 24 10:46:30 2008 Subject: SmartArray E200i (ML350) ''ciss0: ADAPTER HEARTBEAT FAILED'' Message-ID: Hi, A few weeks ago we experienced the same "ciss0: ADAPTER HEARTBEAT FAILED" problem on a dl380 g5 with the HP SmartArray E200 and also the HP SmartArray P400 on FreeBSD 6.2. After booting in verbose mode I saw in dmesg an: ciss0: command status 0x1 (target status) scsi status 0x2 (probe1:ciss0:0:1:0): error 22 (probe1:ciss0:0:1:0): Unretryable Error That was with firmware version 1.20. After upgrading to firmware version 1.78 this error message was gone. I am still not sure what caused this heartbeat failed problem, but I'm running a stresstest for a week now on a michine with and without the latest firmware. Both have an E200 controller. I hope the machine without the latest firmware will have the same heartbeat failed problem soon, and the one with the latest firmware will keep running. That would prove that the firmware update was indeed the solution. My problem at the moment, because we have multiple servers using the same configuration is to find out what firmware version the Smart Array has without having to reboot the machine. I've tried several methods, but none show the firmware version (except rebooting, but that's not really an option). I've found the tool hpacucli, which should do the trick, but I can't get version 7.50 running on FreeBSD 6.2. Has anyone got any ideas how I could handle this problem? Greetings, Blikjeham. From matsuyama at qac.jp Wed Sep 24 12:09:07 2008 From: matsuyama at qac.jp (Junichi Matsuyama) Date: Wed Sep 24 12:09:13 2008 Subject: SmartArray E200i (ML350) ''ciss0: ADAPTER HEARTBEAT FAILED'' In-Reply-To: Message-ID: I don't know how to get firmware version without rebooting... After updating some DL360 G5 servers with Firmware Maintenance CD v8.10 about 2 months ago, "HEARTBEAT FAILED" problem has not been reproduced. -- Junichi Matsuyama From jendries at planetpyramid.com Wed Sep 24 16:11:14 2008 From: jendries at planetpyramid.com (Josh Endries) Date: Wed Sep 24 16:11:17 2008 Subject: SmartArray E200i (ML350) ''ciss0: ADAPTER HEARTBEAT FAILED'' In-Reply-To: References: Message-ID: Hello, Sorry, ProLiant noob here, and sorry to hijack this thread. :) Simple question: I have a DL360 G5 and would like to test out the upgrade. How exactly do you update the firmware for these? I poked around the web site and saw a maintenance CD, and looked at the software for Solaris, though I imagine you could download firmware from any OS. Do you use the USB Key thingy and boot to that, or how is it done with FreeBSD? P.S. I'm not sure it's applicable but in the ciss man page on 7.0 there is reference to a sysctl that may prevent this heartbeat problem (though a firmware upgrade is probably better if possible). Thanks, Josh From ulf at Alameda.net Wed Sep 24 17:37:26 2008 From: ulf at Alameda.net (Ulf Zimmermann) Date: Wed Sep 24 17:37:29 2008 Subject: SmartArray E200i (ML350) ''ciss0: ADAPTER HEARTBEAT FAILED'' In-Reply-To: References: Message-ID: <20080924171923.GJ68150@evil.alameda.net> On Wed, Sep 24, 2008 at 11:59:04AM -0400, Josh Endries wrote: > Hello, > > Sorry, ProLiant noob here, and sorry to hijack this thread. :) Simple > question: I have a DL360 G5 and would like to test out the upgrade. How > exactly do you update the firmware for these? I poked around the web > site and saw a maintenance CD, and looked at the software for Solaris, > though I imagine you could download firmware from any OS. Do you use the > USB Key thingy and boot to that, or how is it done with FreeBSD? > > P.S. I'm not sure it's applicable but in the ciss man page on 7.0 there > is reference to a sysctl that may prevent this heartbeat problem (though > a firmware upgrade is probably better if possible). An easy solution is to download R.I.P. (Rescue is possible) which can be booted via PXE, then use the Linux online upgrade .scexe to flash the firmware. The firmware cd is just a linux booted from CD which has that same .scexe it executes. -- Regards, Ulf. --------------------------------------------------------------------- Ulf Zimmermann, 1525 Pacific Ave., Alameda, CA-94501, #: 510-865-0204 You can find my resume at: http://www.Alameda.net/~ulf/resume.html From sutter.cane at rx78.org Wed Sep 24 20:04:02 2008 From: sutter.cane at rx78.org (Akira Norimaki) Date: Wed Sep 24 20:04:10 2008 Subject: SmartArray E200i (ML350) ''ciss0: ADAPTER HEARTBEAT FAILED'' In-Reply-To: References: Message-ID: <48DA9D2F.8030404@rx78.org> Josh Endries ha scritto: > P.S. I'm not sure it's applicable but in the ciss man page on 7.0 there > is reference to a sysctl that may prevent this heartbeat problem http://freshbsd.org/?branch=&committer=iwasaki&module=&project= Add `hw.ciss.nop_message_heartbeat' tunable (default disabled) for NOP-message polling in ciss_periodic(). Note that setting the tunable to non-zero can be workaround only for `ADAPTER HEARTBEAT FAILED' problem, and may freeze the system w/o the problem. The note frightens me a bit. And then, which non-zero value? How much should the right value be? :) Bye, Akira From blikjeham at vierkleurenspreeuw.nl Thu Sep 25 06:11:49 2008 From: blikjeham at vierkleurenspreeuw.nl (blikjeham) Date: Thu Sep 25 06:11:59 2008 Subject: SmartArray E200i (ML350) ''ciss0: ADAPTER HEARTBEAT FAILED'' In-Reply-To: <48DA9D2F.8030404@rx78.org> References: <48DA9D2F.8030404@rx78.org> Message-ID: <48DB4355.9090609@vierkleurenspreeuw.nl> Akira Norimaki wrote: > Josh Endries ha scritto: > >> P.S. I'm not sure it's applicable but in the ciss man page on 7.0 there >> is reference to a sysctl that may prevent this heartbeat problem >> > > http://freshbsd.org/?branch=&committer=iwasaki&module=&project= > > Add `hw.ciss.nop_message_heartbeat' tunable (default disabled) for > NOP-message polling in ciss_periodic(). > Note that setting the tunable to non-zero can be workaround only for > `ADAPTER HEARTBEAT FAILED' problem, and may freeze the system w/o > the problem. > > I've tried that solution, but something strange occured. On the same machine that first displayed the ADAPTER HEARTBEAT FAILED problem, it froze with this 'workaround'. Without it, it didn't freeze. That's the exact opposite of the expected behavior. On the console it displayed: ciss0: ADAPTER HEARTBEAT FAILED during initialisation. > The note frightens me a bit. And then, which non-zero value? How much > should the right value be? :) > > Bye, > Akira > You can set it to any non-zero value, but one would suffice. It's just a boolean you set in /boot/loader.conf. The solution suggested to me was to upgrade the firmware. Now I'm still looking for a way to review what firmware version I got, without having to reboot the system. BlikjeHam From jendries at planetpyramid.com Sun Sep 28 20:35:02 2008 From: jendries at planetpyramid.com (Josh Endries) Date: Sun Sep 28 20:35:08 2008 Subject: Question about management utilities (hpacucli, camcontrol, etc.) Message-ID: Hello, I'm curious what you all use (if anything) to monitor/tune your disk arrays. I've found hpacucli and hpasm, but the former requires compat4x *and* 5x (?) and the latter only works on 6.2 according to the page I found. I can get a brief status from camcontrol (4 characters), is there anything better? Is there any way to en/disable or list write caching or other options? camcontrol modepage didn't work for me. I have a DL360 G5 with P400i (and a couple other DL models). Josh From jendries at planetpyramid.com Mon Sep 29 15:42:07 2008 From: jendries at planetpyramid.com (Josh Endries) Date: Mon Sep 29 15:42:14 2008 Subject: Question about management utilities (hpacucli, camcontrol, etc.) In-Reply-To: <48DFEE0B.3020401@FreeBSD.org> References: <48DFEE0B.3020401@FreeBSD.org> Message-ID: Hello, I tried impitool, looks neat, but I can't figure out how to get disk information out of it (if possible). root@hathor# ipmitool -I open sdr list all UID Light | 0 unspecified | ok Int. Health LED | 0 unspecified | ok Ext. Health LED | 0 unspecified | ok Power Supply 1 | 0 unspecified | nc Power Supply 2 | 0 unspecified | nc Power Supplies | 0 unspecified | nc VRM 1 | 0 unspecified | cr VRM 2 | 0 unspecified | cr Fan Block 1 | 34.89 unspecifi | nc Fan Block 2 | 29.79 unspecifi | nc Fan Block 3 | 34.89 unspecifi | nc Fan Blocks | 0 unspecified | nc Temp 1 | 42 degrees C | ok Temp 2 | 19 degrees C | ok Temp 3 | 30 degrees C | ok Temp 4 | 30 degrees C | ok Temp 5 | 32 degrees C | ok Temp 6 | 30 degrees C | ok Temp 7 | 30 degrees C | ok Power Meter | 270 Watts | cr I tried a bunch of other commands, such as 'ipmitool -I open sdr type "Drive Slot / Bay"', using -d 0 and 2 (I have ipmi0 and ipmi2 in /dev), but they didn't seem to get me any disk information. Could it be that my P400i doesn't hook into IPMI, or maybe I'm just not grokking ipmitool enough yet. Still a neat utility. Thanks, Josh