From: don schad Newsgroups: comp.parallel.mpi Subject: Linux LAM6.1 hboot timeout error Date: 28 Sep 1998 21:51:02 GMT Organization: Frontier Internet Rochester N.Y. (716)-777-SURF Message-Id: <6up0c6$1foe$1@node17.cwnet.frontiernet.net> User-Agent: tin/pre-1.4-980226 (UNIX) (AIX/4-2) Hi, I am having trouble trying to get LAM 6.1 to lamboot my multi-computer (a Linux 2cpu PPro running Red Hat 5.1, and 1 cpu P-MMX running Slackware) properly. I have 2 nodes (osiris(P-MMX)), breca), and when I use osiris as the origin node LAM bombs out and returns a timeout error. I can do a recon, which indicates that all the nodes are up and running, but when i actually try to boot the system it doesnt work: osiris:/usr/local/linuxshare/tests%recon -v pp2hosts recon: testing n0 (breca) recon: testing n1 (osiris) osiris:/usr/local/linuxshare/tests%lamboot -v pp2hosts LAM 6.1 - Ohio Supercomputer Center hboot n0 (breca)... lamboot: hboot failed on n0 (breca) lamboot: Connection timed out wipe ... tkill n0 (breca)... osiris:/usr/local/linuxshare/tests% When I use breca as the origin computer, everything works fine: breca:/usr/local/linuxshare/tests>recon -v pp2hosts recon: testing n0 (breca) recon: testing n1 (osiris) breca:/usr/local/linuxshare/tests>lamboot -v pp2hosts LAM 6.1 - Ohio Supercomputer Center hboot n0 (breca)... hboot n1 (osiris)... topology done breca:/usr/local/linuxshare/tests> I can start a LAM using osiris and another computer with either of these as the origin system, and it works fine either way. rsh does work from both computers, so I don't think that it is a permission problem, and when I rsh from osiris I can find hboot in my path). I would appreciate any thoughts/help very much. Thanks, don Hydroqual, Inc. -- ------------------------------------------------------- Brought to you sloooooooowly by Frontier Communications, the Intermittent Service Provider.