Re: [SSI-users] Problem booting cluster
Brought to you by:
brucewalker,
rogertsang
From: Jose A. R. <jos...@ac...> - 2005-01-10 14:26:22
|
Well, I tried the 1.2.0 release and I get more or less the same behaviour. I boot 32 nodes with no problems, but the number 33 hangs when executing the init.d scripts. At the same time the master node seems to hang when you try to access the filesystem. I used the "btt" command to get a stack trace of a few processes: (node1) - lock_dir, text.lock.namei, __down_failed, __down (other node) - ncscli_wait, cli_icsnsc_rcall, nsc_rcall, cfs_proc_lookup My bet is that the CFS has some type of deadlock... I would like to activate debug for CFS, but when I try to make a full kernel I got this: [root@ferrer-1 openssi]# make fullkern Cleaning sandboxes Copying ../linux to ../linux-ssi Applying UML patch to ../linux-ssi Copying CI code into ../linux-ssi Copying OpenSSI code into ../linux-ssi Applying i386 patches to ../linux-ssi >>> Applying kdb-i386-ssi.patch make: *** [fullkern] Error 1 Directories ci, openssi and srpms are cvs updated. "linux" dir points to linux-2.4.22-1.2199.nptl_ssi_6. Any ideas? > If what you want is the ability to generate netdumps, it might still > work with the 1.2.0 kernel. > > If you want KDB, edit the kernel args in grub.conf to delete the > 'panic=15' and add 'kdb=on'. If you're using PXE or Etherboot, you'll > also need to make a similar change to /tftpboot/pxelinux.cfs/default and > the mkelf-linux command at the bottom of /sbin/ssi-ksync-network, > respectively. > > I seem to remember that 'kdb=on' did not work very well with the kernel > for OpenSSI 1.0.0. The 1.2.0 release has a newer kernel, so it may work, > but if it doesn't, you'll need to rebuild the kernel with the > CONFIG_KDB_OFF option disabled. > > Let us know how this goes, > > Brian Jose ____________________________________________________________________________ Jose A. Rodriguez OOO Universitat Politecnica de Catalunya (UPC) jo...@ac... OOO Departament d'Arquitectura de Computadors Tel. 16990 OOO -*- LCAC -*- UPC |