Squid first saves object in memory. Then swapout object to cache. As usual:
# MEMORY CACHE OPTIONS
#
-----------------------------------------------------------------------------
# TAG: cache_mem (bytes)
# NOTE: THIS PARAMETER DOES NOT SPECIFY THE MAXIMUM PROCESS SIZE.
# IT ONLY PLACES A LIMIT ON HOW MUCH ADDITIONAL MEMORY SQUID WILL
# USE AS A MEMORY CACHE OF OBJECTS. SQUID USES MEMORY FOR OTHER
# THINGS AS WELL. SEE THE SQUID FAQ SECTION 8 FOR DETAILS.
#
# 'cache_mem' specifies the ideal amount of memory to be used
# for:
# * In-Transit objects
# * Hot Objects
# * Negative-Cached objects
#
# Data for these objects are stored in 4 KB blocks. This
# parameter specifies the ideal upper limit on the total size of
# 4 KB blocks allocated. In-Transit objects take the highest
# priority.
#
# In-transit objects have priority over the others. When
# additional space is needed for incoming data, negative-cached
# and hot objects will be released. In other words, the
# negative-cached and hot objects will fill up any unused space
# not needed for in-transit objects.
#
# If circumstances require, this limit will be exceeded.
# Specifically, if your incoming request rate requires more than
# 'cache_mem' of memory to hold in-transit objects, Squid will
# exceed this limit to satisfy the new requests. When the load
# decreases, blocks will be freed until the high-water mark is
# reached. Thereafter, blocks will be used to store hot
# objects.
#
# If shared memory caching is enabled, Squid does not use the shared
# cache space for in-transit objects, but they still consume as much
# local memory as they need. For more details about the shared memory
# cache, see memory_cache_shared.
#Default:
# cache_mem 256 MB
# TAG: maximum_object_size_in_memory (bytes)
# Objects greater than this size will not be attempted to kept in
# the memory cache. This should be set high enough to keep objects
# accessed frequently in memory to improve performance whilst low
# enough to keep larger objects from hoarding cache_mem.
#Default:
# maximum_object_size_in_memory 512 KB
# TAG: memory_cache_shared on|off
# Controls whether the memory cache is shared among SMP workers.
#
# The shared memory cache is meant to occupy cache_mem bytes and replace
# the non-shared memory cache, although some entities may still be
# cached locally by workers for now (e.g., internal and in-transit
# objects may be served from a local memory cache even if shared memory
# caching is enabled).
#
# By default, the memory cache is shared if and only if all of the
# following conditions are satisfied: Squid runs in SMP mode with
# multiple workers, cache_mem is positive, and Squid environment
# supports required IPC primitives (e.g., POSIX shared memory segments
# and GCC-style atomic operations).
#
# To avoid blocking locks, shared memory uses opportunistic algorithms
# that do not guarantee that every cachable entity that could have been
# shared among SMP workers will actually be shared.
#
# Currently, entities exceeding 32KB in size cannot be shared.
#Default:
# "on" where supported if doing memory caching with multiple SMP workers.
# TAG: memory_cache_mode
# Controls which objects to keep in the memory cache (cache_mem)
#
# always Keep most recently fetched objects in memory (default)
#
# disk Only disk cache hits are kept in memory, which means
# an object must first be cached on disk and then hit
# a second time before cached in memory.
#
# network Only objects fetched from network is kept in memory
#Default:
# Keep the most recently fetched objects in memory
This is no memory leaking, but normal cache behaviour. As documented.
You can play around with range_offset_limit and quick_abort_min parameters.
Or try to no cache this FTP with ACL.
Usually, when suggests memory leaking, this often OS issue. Not Squid.
11.02.15 16:02, Silamael пишет:
Hi,
One of our customers does constantly mirroring of some FTP directories
and noticed a huge memory consumption of Squid. As far as I can see with
squidclient mgr:mem, the 2K buffers are constantly increasing if Squid
is processing FTP requests like wget ftp://some.server/pub/ or wget -m
ftp://...
In my tests the first variant resulted in an increase of 29 kb per request.
If I directly request a certain file, the increase of the 2K buffers
seems not to happen.
I have reproduced this behaviour with both squid 3.2.13 and also 3.4.6
and 3.4.11.
Greeting,
Matthias
_______________________________________________
squid-users mailing list
squid-users@xxxxxxxxxxxxxxxxxxxxx
http://lists.squid-cache.org/listinfo/squid-users
_______________________________________________
squid-users mailing list
squid-users@xxxxxxxxxxxxxxxxxxxxx
http://lists.squid-cache.org/listinfo/squid-users
