Re: [phenixbb] exporting results, reduce

I'm finding that Reduce in dev-1616 is taking from 10 to 30 seconds to run (approximately). Does that sound about right? If not, do you have any suggestions for speeding it up? That might not seem like a long time, but I'm running it on the entire PDB, so every second added to the average processing time adds a whole day to the complete run. (My system is 32-bit Ubuntu 12.04 on Intel Xeon CPU X5470 @ 3.33GHz × 8 with 16GB RAM.) Thanks. :) Eric On Fri, Jan 24, 2014 at 12:00 PM, Pavel Afonine wrote:

Hello,

usually Reduce takes from a fraction of a second to seconds to run (provided I understand correctly that Jeff fixed all problems related to "reduce runs forever"). I wonder how PDB file should look like so Reduce runs hours. Could you please send me PDB file (off mailing list, to my email directly)?

Thanks, Pavel

On 1/24/14, 8:44 AM, jp d wrote:

hi, replying for the sake of posterity. the slow export/reduce seems to be related to a malformed input model. the segids are ok but there are some different segids that have the same chain. setting the chain to empty for everything resolved the issue. Using the malformed chain model took 35 hours (if it finished). Using the no chain model took 5 hours. thanks jpd

------------------------------ *From:* jp d *To:* PHENIX user mailing list *Sent:* Friday, January 17, 2014 9:53 AM *Subject:* [phenixbb] exporting results, reduce

hi, we have a large complex at low resolution (7A) and just want to see general conformational changes. Occasionally phenix gets stuck at 'Exporting Results' and top shows 'reduce' running and using lots of RAM . so much RAM that it starts using swap. is there a way to bypass or limit the reduce step. thanks jpd

_______________________________________________ phenixbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/phenixbb

_______________________________________________ phenixbb mailing [email protected]http://phenix-online.org/mailman/listinfo/phenixbb

_______________________________________________ phenixbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/phenixbb