On Wed, 2013-04-10 at 09:21 +0200, Fabien Wernli wrote: > Hi, > > On Tue, Apr 09, 2013 at 09:44:17AM -0700, Albert Chu wrote: > > Did you definitely try the "solpayloadsize" workaround? That workaround > > specifically handles this error: > > I did. The result is a (supposedly) endless loop: > > | LOOPSTART | > = Get Authentication Capabilities V20 Request = > = Get Authentication Capabilities V20 Response = > = Open Session Request = > = Open Session Response = > = Rakp Message 1 Request = > = Rakp Message 2 Response = > = Rakp Message 3 Request = > = Rakp Message 4 Response = > = Set Session Privilege Level Request = > = Set Session Privilege Level Response = > = Set Session Privilege Level Request = > = Set Session Privilege Level Response = > = Get Channel Payload Support Request = > = Get Channel Payload Support Response = > = Get Payload Activation Status Request = > = Get Payload Activation Status Response = > = Activate Payload Request = > = Activate Payload Response = > = Deactivate Payload Request = > = Deactivate Payload Response = > = Close Session Request = > = Close Session Response = > | GOTO LOOPSTART |
Hmmm, an endless loop is really bad. The only time libipmiconsole will loop at all is if the remote BMC informs it that it should use an alternate port. This is very rare on BMCs and it's possible there's a bug in the code (I believe ipmitool just assumes the default port all of the time, ignoring if the remote BMC tells it to use a different port.) Could you try both the "solpayloadsize" and the "solport" workarounds at the same time (-W solpayloadsize,solport). If that doesn't work, can you send me a --debug output when just using "solpayloadsize" (just give me a few loops of the --debug output). > When interrupted, and retried several times, the BMC seems to go on strike > for a minute or two: > > | [error received]: BMC Error This isn't too unusual. Many BMC's will intentionally disable connections if they believe there are too many connections going on too quickly (i.e. someone is trying to do a brute force password attempt). Al -- Albert Chu ch...@llnl.gov Computer Scientist High Performance Systems Division Lawrence Livermore National Laboratory _______________________________________________ Freeipmi-users mailing list Freeipmi-users@gnu.org https://lists.gnu.org/mailman/listinfo/freeipmi-users
