US 7,321,982 B2 | ||
System and method for takeover of partner resources in conjunction with coredump | ||
Susan M. Coatney, Cupertino, Calif. (US); John Lloyd, Roseville, Calif. (US); Jeffrey S. Kimmel, Chapel Hill, N.C. (US); Brian Parkison, Felton, Calif. (US); and David Brittain Bolen, Durham, N.C. (US) | ||
Assigned to Network Appliance, Inc., Sunnyvale, Calif. (US) | ||
Filed on Jan. 26, 2004, as Appl. No. 10/764,809. | ||
Prior Publication US 2005/0177770 A1, Aug. 11, 2005 | ||
Int. Cl. G06F 11/00 (2006.01) |
U.S. Cl. 714—4 [714/11; 714/13] | 28 Claims |
1. A method for taking over a failed filer owning disks that store file service data and owning at least one disk that is
free of file service data by a clustered partner filer, the failed filer being adapted to perform a coredump in which, in
an event of failure, memory contents of the failed filer are transferred to a disk, comprising:
changing, by the failed filer, a coredump attribute on the disk that is free of file service data (the “coredump disk”) from
a non-coredump state to a coredump state and maintaining the coredump attribute on other disks owned by the failed filer in
a non-coredump state;
writing the memory contents to the coredump disk;
identifying, by the clustered partner filer, the coredump attribute of the other disks and taking ownership of the other disks
while allowing the failed filer to maintain ownership of the coredump disk;
upon completion of the writing of the memory contents, changing the coredump attribute to a non-coredump state; and
upon identification of the non-coredump state in the coredump attribute of the coredump disk, taking ownership, by the clustered
partner filer, of the coredump disk.
|