0000238: Incompatibility between new sp2+hackelec and cluster

Notes
(0000238) momeara (Attentive Developer) 2013-04-02 12:13	Adding -corrections:hbond_sp2_correction -score:weights sp2_correction.wts to the cluster integration test does not segfault. Doonam, can you please either add or link all the input files so that I can reproduce the segfault?

(0000239) rmoretti (Attentive Developer) 2013-04-02 12:51	It looks like it's a issue with the Dunbrack library. -corrections:hbond_sp2_correction turns on the dun10 library. If for some reason you had a corrupt dun10 library, that could account for the crash with the flag, but no crash without the flag (i.e. using dun02). Make sure your database is up-to-date and matches the files on the server (i.e. svn status doesn't report any changed), and try deleting the generated dun10 binary (e.g. rosetta_database/rotamer/ExtendedOpt1-5/Dunbrack10.lib.bin), if any. (Keep a copy of the binary for later post analysis - we should give an error message if the binary is corrupt, rather than a segfault.) The other thing is if you're running this on multiple processors/threads, you may be encountering write collisions with generating the dun10 binary for the first time. Try doing a single-processor single-threaded score with the -corrections:hbond_sp2_correction to pre-generate the binary prior to launching a multiprocessor/multithreaded cluster run.

(0000240) kimdn (Developer) 2013-04-02 13:40	After reading rocco's note, I erased "rosetta_database/rotamer/ExtendedOpt1-5/Dunbrack10.lib.bin" and ran cluster again and this time it ran well like below. Since 'Dunbrack10.lib.bin_led_seg_fault_compiled_at_54637' is too big, I copy the file into Kuhlman lab's internal site. (@ contador /home/kimdn/scr/data/do_not_erase/Dunbrack10.lib.bin_led_seg_fault_compiled_at_54637) "core.pack.task: Packer task: initialize from command line() Warning: Unable to locate database file rotamer/ExtendedOpt1-5/Dunbrack10.lib.bin core.pack.dunbrack: Reading Dunbrack Libraries core.pack.dunbrack: Reading /nas02/home/k/i/kimdn/rosetta/rosetta_database/rotamer/ExtendedOpt1-5/cys.bbdep.rotamers.lib ... core.pack.dunbrack: Reading /nas02/home/k/i/kimdn/rosetta/rosetta_database/rotamer/ExtendedOpt1-5/asp.bbdep.rotamers.lib core.pack.dunbrack: Reading /nas02/home/k/i/kimdn/rosetta/rosetta_database/rotamer/ExtendedOpt1-5/asp.bbdep.densities.lib core.pack.dunbrack: Reading /nas02/home/k/i/kimdn/rosetta/rosetta_database/rotamer/ExtendedOpt1-5/asp.bbind.chi2.Definitions.lib Creating tricubic splines for the backbone-dependent non-rotameric chi scores Finished creating tricubic splines core.pack.dunbrack: Reading /nas02/home/k/i/kimdn/rosetta/rosetta_database/rotamer/ExtendedOpt1-5/glu.bbdep.rotamers.lib core.pack.dunbrack: Reading /nas02/home/k/i/kimdn/rosetta/rosetta_database/rotamer/ExtendedOpt1-5/glu.bbdep.densities.lib core.pack.dunbrack: Reading /nas02/home/k/i/kimdn/rosetta/rosetta_database/rotamer/ExtendedOpt1-5/glu.bbind.chi3.Definitions.lib Creating tricubic splines for the backbone-dependent non-rotameric chi scores Finished creating tricubic splines core.pack.dunbrack: Reading /nas02/home/k/i/kimdn/rosetta/rosetta_database/rotamer/ExtendedOpt1-5/phe.bbdep.rotamers.lib core.pack.dunbrack: Reading /nas02/home/k/i/kimdn/rosetta/rosetta_database/rotamer/ExtendedOpt1-5/phe.bbdep.densities.lib core.pack.dunbrack: Reading /nas02/home/k/i/kimdn/rosetta/rosetta_database/rotamer/ExtendedOpt1-5/phe.bbind.chi2.Definitions.lib ... core.pack.dunbrack: Reading /nas02/home/k/i/kimdn/rosetta/rosetta_database/rotamer/ExtendedOpt1-5/asn.bbdep.densities.lib core.pack.dunbrack: Reading /nas02/home/k/i/kimdn/rosetta/rosetta_database/rotamer/ExtendedOpt1-5/asn.bbind.chi2.Definitions.lib Creating tricubic splines for the backbone-dependent non-rotameric chi scores Finished creating tricubic splines ... core.pack.dunbrack: Dunbrack 2010 library took 35.51 seconds to load from ASCII core.pack.dunbrack: Memory usage: core.pack.dunbrack: CYS with 187664 bytes ... core.pack.dunbrack: TYR with 5789104 bytes core.pack.dunbrack: Total memory on Dunbrack Libraries: 93727608 bytes. Warning: Unable to locate database file rotamer/ExtendedOpt1-5/Dunbrack10.lib.bin Warning: Unable to locate database file rotamer/ExtendedOpt1-5/Dunbrack10.lib.bin core.pack.dunbrack: Opening file /nas02/home/k/i/kimdn/rosetta/rosetta_database/dun10XtCuTa for output. protocols.cluster: RESCORING: S_00000345_from_2_1_0001.pdb protocols.cluster: Adding struc: 2.122e-314 core.pack.task: Packer task: initialize from command line() protocols.cluster: RESCORING: S_00000345_from_2_1_0002.pdb"

(0000241) kimdn (Developer) 2013-04-02 13:45	After reading rocco's note, I erased "rosetta_database/rotamer/ExtendedOpt1-5/Dunbrack10.lib.bin" and ran cluster again and this time it ran well. Since 'Dunbrack10.lib.bin_led_seg_fault_compiled_at_54637' is too big, I copy the file into Kuhlman lab's internal site. (@ contador /home/kimdn/scr/data/do_not_erase/Dunbrack10.lib.bin_led_seg_fault_compiled_at_54637)

Issue History
Date Modified	Username	Field	Change
2013-04-02 11:39	kimdn	New Issue
2013-04-02 11:39	kimdn	File Added: flag_cluster
2013-04-02 12:11	momeara	Assigned To	=> momeara
2013-04-02 12:11	momeara	Status	new => assigned
2013-04-02 12:13	momeara	Note Added: 0000238
2013-04-02 12:51	rmoretti	Note Added: 0000239
2013-04-02 13:40	kimdn	Note Added: 0000240
2013-04-02 13:44	kimdn	Description Updated	View Revisions
2013-04-02 13:45	kimdn	Fixed in SVN Version	=> 54637
2013-04-02 13:45	kimdn	Note Added: 0000241
2013-04-02 13:45	kimdn	Status	assigned => closed
2013-04-02 13:45	kimdn	Resolution	open => fixed
2013-04-02 13:45	kimdn	Fixed in Version	=> Trunk
2013-04-02 13:54	momeara	Status	closed => feedback
2013-04-02 13:54	momeara	Resolution	fixed => reopened

Relationships