[linux] paralelne spustanie procesov (kompresia logov)

[Peter Viskup]

make argument -j obmedzuje iba max. mnozstvo beziacich procesov.
Strazi si make aj rozlozenie procesov cez co najviac CPU?

--
Peter Viskup

Lubomir Host wrote:
> On Mon, Feb 01, 2010 at 09:47:54PM +0100, Matus UHLAR - fantomas wrote:
>   
>>> On Mon, Feb 01, 2010 at 03:38:51AM +0100, Matus UHLAR - fantomas wrote:
>>>       
>>>> poznate niekto aplikaciu ktora by mi umoznila zadefinovat niekolko procesov
>>>> ktore by sa mali spustit, pricom paralelne by malo bezat vzdy max. X
>>>> procesov?
>>>>
>>>> rad by som komprimoval dost velke logy paralelne na kazdom procesore jeden
>>>>         
>> On 01.02.10 08:13, Vitezslav Samel wrote:
>>     
>>>     make  a jeho parametr -j
>>>       
>> mas nejaky skript ktory za tymto ucelom pouzivas?
>> Lebo ja sa bojim ze toto by bolo predsa len trochu komplikovane pri sposobe
>> ktorym to pouzivame my.
>>     
>
> Skus tento paralell.sh
>
> A neslo by make-u podhodit pravidla shruba taketo (neskusane):
>
> SOURCE  = $(shell find /var/log/  -type f -name '*.log' -print)
> COMPRESSED = $(patsub $(SOURCE),%.log,%.log.bz2)
> all: $(COMPRESSED)
> %.log.bz2: %.log
>     bzi2p --best %<
>
> Daj vediet ako si to nakoniec vyriesil.
>
> rajo
>
>   
> ------------------------------------------------------------------------
>
> _______________________________________________
> https://lists.linux.sk/mailman/listinfo/linux
> Meta FAQ: http://www.sklug.sk/lists/linux/metafaq.html