I think you need to do:
rc_toggle_basic panic "this is my panic message..."
in order to force a core dump at the same moment. This is for strong hearts only !
eT.
--- Brian Tao taob@risc.org wrote:
On Fri, 21 Jul 2000, Bennett Todd wrote:
At 5 minutes, that means Netapps are no longer booting
significantly
faster than generic servers.
I just tried a reboot on one of our pre-production F740's, andthe NFS client (a Solaris machine) saw an NFS outage of 1m16s. Granted, this doesn't take into account a savecore, and I'm not sure how to simulate that. It would be nice to have a firmware setting to disable the boot up memory scan and save another 10 or 15 seconds. -- Brian Tao (BT300, taob@risc.org) "Though this be madness, yet there is method in't"
===== Yours, Eyal Traitel eTraitel@yahoo.com, Home: 972-3-5290415 (Tel Aviv) *** eTraitel - it's the new eBuzzword ! ***
__________________________________________________ Do You Yahoo!? Get Yahoo! Mail � Free email you can access from anywhere! http://mail.yahoo.com/
etraitel@yahoo.com (Eyal Traitel) writes:
On Fri, 21 Jul 2000, Bennett Todd wrote: I just tried a reboot on one of our pre-production F740's, and the NFS client (a Solaris machine) saw an NFS outage of 1m16s. Granted, this doesn't take into account a savecore, and I'm not sure how to simulate that. It would be nice to have a firmware setting to disable the boot up memory scan and save another 10 or 15 seconds.
I think you need to do:
rc_toggle_basic panic "this is my panic message..."
in order to force a core dump at the same moment. This is for strong hearts only !
"reboot -d" is the officially sanctioned way to do this, surely?
Chris Thompson University of Cambridge Computing Service, Email: cet1@ucs.cam.ac.uk New Museums Site, Cambridge CB2 3QG, Phone: +44 1223 334715 United Kingdom.
On Sat, 22 Jul 2000, Eyal Traitel wrote:
I think you need to do:
rc_toggle_basic panic "this is my panic message..."
in order to force a core dump at the same moment. This is for strong hearts only !
Yep, I discovered that command shortly after my previous message. Tried it twice. First time, 3m20s from panic to completed reboot. The filer was panicked while a bunch of writes were going on, so this should account for the NVRAM replay:
Sat Jul 22 02:17:43 EDT [rc]: GbE-e8: Link up Sat Jul 22 02:17:43 EDT [rc]: saving 45M to /etc/crash/core.0.nz ("user panic: on release NetApp Release 5.3.6R1") Sat Jul 22 02:17:47 EDT [rc]: relog syslog PANIC: user panic: on release NetApp Release 5.3.6R1 on Sat Jul 22 06:14:30 2000 Sat Jul 22 02:17:47 EDT [rc]: NetApp Release 5.3.6R1 boot complete. Last disk update written at Sat Jul 22 02:14:27 EDT 2000 Sat Jul 22 02:17:47 EDT [mini_core_admin]: generating /etc/crash/mini-core.0
I tried it again though (since I wasn't playing close enough attention the first time), and it took almost 15 minutes (!) for writes to resume on the NFS client:
Sat Jul 22 02:39:01 EDT [rc]: GbE-e8: Link up Sat Jul 22 02:39:01 EDT [rc]: saving 18M to /etc/crash/core.1.nz ("user panic: on release NetApp Release 5.3.6R1") Sat Jul 22 02:39:03 EDT [rc]: relog syslog PANIC: user panic: on release NetApp Release 5.3.6R1 on Sat Jul 22 06:24:15 2000 Sat Jul 22 02:39:03 EDT [rc]: NetApp Release 5.3.6R1 boot complete. Last disk update written at Sat Jul 22 02:24:10 EDT 2000 Sat Jul 22 02:39:03 EDT [mini_core_admin]: generating /etc/crash/mini-core.1 Sat Jul 22 02:39:23 EDT [mini_core_admin]: compressed core '/etc/crash/core.1.nz' has errors.
I did delete the contents of /etc/crash before the second panic, there was lots of disk space (basically an empty 5x36GB filesystem), and the filer did not reboot multiple times. I wasn't at the console to watch the boot messages go by. Anyone seen this?
-
Brian Tao -
Chris Thompson -
Eyal Traitel