WARNING: Currently does not work properly in combination with
rate-control.
I'm caching the files taken from a mounted network file system to a
shared local directory. Multiple instances can share the same cache
directory, but it's recommended to use slightly different values for the
property cleanup-high-watermark. I'm using the directory access time to
know when the cache usage changed and keep an estimation of the cache
usage for statistics.
I'm creating a unique thread to do the file copying block by block,
for all files to be copied to the cache. Using a thread instead of a
reactor.callLater 'loop' allow for higher copy throughput and do not slow
down the mail loop when lots of files are copied at the same time.
Simulations with real request logs show that using a thread gives better
results than the equivalent asynchronous implementation.
|
__init__(self,
args)
x.__init__(...) initializes x; see help(type(x)) for signature |
source code
|
|
|
|
|
|
|
|
|
|
FilePath
|
|
|
getLogName(self,
path,
id=None)
Returns a log name for a path, shortened to a maximum size specified
by the global variable MAX_LOGNAME_SIZE. |
source code
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Inherited from plugs.base.ComponentPlug :
restart
Inherited from object :
__delattr__ ,
__format__ ,
__getattribute__ ,
__hash__ ,
__new__ ,
__reduce__ ,
__reduce_ex__ ,
__repr__ ,
__setattr__ ,
__sizeof__ ,
__str__ ,
__subclasshook__
Inherited from extern.log.log.Loggable :
__providedBy__ ,
__provides__ ,
debug ,
doLog ,
error ,
info ,
log ,
logFunction ,
logObjectName ,
warning ,
warningFailure ,
writeMarker
|