Re: [squid-users] Squid 3.3 is very aggressive with memory

On 22/12/13 09:37, Nathan Hoad wrote:
> Sure thing. I've put the squid.conf and info manager page onto my
> server, to save sending everyone a very large email.
>
> anonymised+trimmed squid.conf: http://getoffmalawn.com/static/squid.conf
> info manager page: http://getoffmalawn.com/static/squid-manager-info.out
>
> [root_at_host ~]# free -m
> total used free shared buffers cached
> Mem: 3040 2916 124 0 33 355
> -/+ buffers/cache: 2526 514
> Swap: 645 342 302

OK so it seems like the basic issue is probably not in the level of FD
or something similar.
I do not know why the usage of ram is so high but the basic test I would
do is:
remove any cache_dir from squid
remove any memory cache from squid
make sure that the swap is at 0 state.
make sure what process do run on this machine..

I have a tiny proxy with 1GB ram and it works with no problems that I
know of yet with 3.4.X.

Since it's a 32bit machine and has couple things I have a spec file that
worked fine for it on the latest build of squid.

If you do not use authentication remove any of the non-relevant build
instructions..

Also if you can get into a position where all of your machine details
are known such as "what each process uses??" you can answer yourself more.

ps avx -L |grep squid
can give more but if you do have top output for all squid process and
subprocess it will be nice.
as you can see these processes do use an amount of ram but if make sure
that the ram is "slipping" from one way or another and in not accounted
in top we have a more solid ground on that.

What would "http://squid_ip:port/squid-internal-mgr/mem"
Would give us?

it has a lot of details that can be analyzed..

Can you by any chance describe the size of the network?
We are not talking about a home network..
29 clients? ip addresses?

I would try to run the proxy with no cache at all for about 24\48 hours
to see if there is a peek in the FD usage or any load on the service
which can describe what you are talking about.

There is always the possibility for a memory leak and it's not always
related directly to squid..

What are the network requirements? Are you using WCCP(for example)?

All The Bests,
Eliezer

>
> For reference, from top:
>
> 17715 squid 15 0 788m 766m 5528 D 19.6 25.2 18:12.23 (squid-1)
> -f /etc/squid/squid.conf
> 13183 squid 15 0 624m 600m 5444 S 0.0 19.7 14:57.00 (squid-1)
> -f /etc/squid/squid.conf2 # second instance, serving transparent
> traffic only, currently inactive
>
> Note that the manager, free and top output are not free my most recent
> run (i.e. they will not match the cache.log I'll mention later). There
> was a bit of an emergency with the last run with full logging, that
> required I kill everything before I had time to gather these pieces of
> information. Running out of RAM does that...
>
> Any directive mentioned in the squid.conf can be disabled, if it will
> help isolation of the leak. I'm also happy to rebuild Squid with any
> other parameters that may be useful. I'm currently building via an RPM
> spec file on a CentOS 5 box, with gcc 4.1.2 (ancient, I know!) and the
> following options:
>
> Squid Cache: Version 3.3.11
> configure options: '--build=i686-redhat-linux-gnu'
> '--host=i686-redhat-linux-gnu' '--target=i586-redhat-linux-gnu'
> '--program-prefix=' '--prefix=/usr' '--exec-prefix=/usr'
> '--bindir=/usr/bin' '--sbindir=/usr/sbin' '--sysconfdir=/etc'
> '--datadir=/usr/share' '--includedir=/usr/include' '--libdir=/usr/lib'
> '--libexecdir=/usr/libexec' '--sharedstatedir=/usr/com'
> '--mandir=/usr/share/man' '--infodir=/usr/share/info'
> '--exec_prefix=/usr' '--libexecdir=/usr/lib/squid'
> '--localstatedir=/var' '--datadir=/usr/share/squid'
> '--sysconfdir=/etc/squid' '--with-logdir=$(localstatedir)/log/squid'
> '--with-pidfile=$(localstatedir)/run/squid.pid'
> '--disable-dependency-tracking' '--enable-arp-acl'
> '--enable-follow-x-forwarded-for' '--enable-auth'
> '--enable-auth-basic=DB,LDAP,MSNT,MSNT-multi-domain,NCSA,NIS,PAM,POP3,RADIUS,SASL,SMB,getpwnam'
> '--enable-auth-ntlm=smb_lm,fake'
> '--enable-auth-digest=file,LDAP,eDirectory' '--enable-auth-negotiate'
> '--enable-external-acl-helpers=ip_user,ldap_group,session,unix_group,wbinfo_group'
> '--enable-cache-digests' '--enable-cachemgr-hostname=localhost'
> '--enable-delay-pools' '--enable-epoll' '--enable-icap-client'
> '--enable-ident-lookups' '--with-large-files'
> '--enable-linux-netfilter' '--enable-referer-log'
> '--enable-removal-policies=heap,lru' '--enable-snmp' '--enable-ssl'
> '--enable-ssl-crtd' '--enable-storeio=aufs,diskd,ufs'
> '--enable-useragent-log' '--enable-wccpv2' '--with-aio'
> '--with-default-user=squid' '--with-dl' '--with-openssl'
> '--with-pthreads' '--disable-ipv6' '--disable-loadable-modules'
> '--disable-eui' 'build_alias=i686-redhat-linux-gnu'
> 'host_alias=i686-redhat-linux-gnu'
> 'target_alias=i586-redhat-linux-gnu' 'CFLAGS=-O2 -g -pipe -Wall
> -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector
> --param=ssp-buffer-size=4 -m32 -march=i586
> -fasynchronous-unwind-tables -fpie' 'LIBS=-lpresenceclient
> -L/usr/local/lib' 'CXXFLAGS=-O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2
> -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m32
> -march=i586 -fasynchronous-unwind-tables -fpie'
> 'PKG_CONFIG_PATH=/usr/lib/pkgconfig:/usr/share/pkgconfig'
> --enable-ltdl-convenience
>
> Thanks,
>
> Nathan.
>
> On Wed, Dec 18, 2013 at 11:38 AM, Eliezer Croitoru <eliezer_at_ngtech.co.il> wrote:
>> OK Nathan,
>>
>> The next steps are squid.conf..
>> Which can clarify couple things.
>> Also you do have the cache-mgr interface in http and in it you have
>> statistics.
>> http://proxy_ip:3128/squid-internal-mgr/info
>> (The above example).
>> It will provide much more data then just like that looking at memory usage.
>> Also please provide "free -m" output.
>>
>> Thanks,
>> Eliezer
>>
>>
>> On 17/12/13 07:24, Nathan Hoad wrote:
>>>
>>> Okay, to follow up. I still cannot reproduce this in a lab
>>> environment, but I have implemented a way of doing what Alex described
>>> on the production machine. I run two instances of Squid with the same
>>> config and switch the transparent proxy out by changing the redirect
>>> rules in iptables. The second instance is running without a cache_dir
>>> though, to prevent the possibility of two instances sharing the same
>>> directory and running amok. If requested, I can create a second
>>> cache_dir for the second instance to mimic the config entirely.
>>>
>>> While running under this configuration, I've confirmed that memory
>>> usage does go up when active, and stays at that level when inactive,
>>> allowing some time for timeouts and whatnot. I'm currently switching
>>> between the two instances every fifteen minutes.
>>>
>>> Here is a link to the memory graph for the entire running time of the
>>> second process, at 1 minute intervals:
>>> http://getoffmalawn.com/static/mem-graph.png. The graph shows memory
>>> use steadily increasing during activity, but remaining reasonably
>>> stable during inactivity.
>>>
>>> Where shall we go from here? Given that I can switch between the
>>> instances, impacting performance on the production box is not of huge
>>> concern now, so I can run the second instance under Valgrind, or bump
>>> up the debug logging, or whatever would be helpful.
>>>
>>> As an aside, I've been reading some of the code pointed at by traces
>>> I've got, and I've stumbled upon the fact that nearly every caller of
>>> StoreEntry::replaceHttpReply will leak HttpReply objects if the
>>> internal mem_obj pointer of a StoreEntry is set to NULL. There's a
>>> critical log message that occurs in this situation which I have not
>>> seen, so I can conclude that this is not the issue I am seeing, but
>>> it's an issue nonetheless. If there's interest, I'll submit a patch
>>> for this issue.
>>>
>>> Many thanks,
>>>
>>> Nathan.
>>> --
>>> Nathan Hoad
>>> Software Developer
>>> www.getoffmalawn.com
>>>
>>>
>>> On Sat, Dec 14, 2013 at 8:11 PM, Nathan Hoad <nathan_at_getoffmalawn.com>
>>> wrote:
>>>>
>>>> On Fri, Dec 13, 2013 at 10:33 PM, Eliezer Croitoru <eliezer_at_ngtech.co.il>
>>>> wrote:
>>>>>
>>>>> Hey Nathan,
>>>>>
>>>>> I am looking for more details on the subject in hand in the shape of:
>>>>> Networking Hardware
>>>>
>>>>
>>>> Straight out of lspci:
>>>>
>>>> 02:00.0 Ethernet controller: Broadcom Corporation NetXtreme BCM5722
>>>> Gigabit Ethernet PCI Express
>>>> 03:01.0 Ethernet controller: Broadcom Corporation NetXtreme BCM5703
>>>> Gigabit Ethernet (rev 10)
>>>>
>>>> Two network cards - one for internal traffic, the other for external.
>>>>
>>>>> Testing Methods
>>>>
>>>>
>>>> - a mixture of direct and intercepted HTTP and HTTPS traffic, hitting
>>>> the configured ICAP server and not.
>>>> - both valid and invalid upstream SSL certificates, hundreds of
>>>> concurrent requests from a single client
>>>> - thrashing Squid with thousands of connections that are aborted
>>>> after 800ms, running for ~30-40 seconds at a time.
>>>> - currently I'm putting the week's access.log through Squid to see if
>>>> that triggers it, for a poor approximation of the traffic.
>>>>
>>>>> Is it a SMP squid setup?
>>>>
>>>> - both SMP (2 workers) and non-SMP.
>>>>
>>>>> In the case you use a 32bit system which is limited to how much ram??(I
>>>>> remember something about a windows nt with 64GB).
>>>>
>>>>
>>>> - This particular host has 3gb of RAM. Previously running a non-SMP
>>>> Squid 3.2.13 instance and according to logs, maxed out at ~500mb of
>>>> resident after running for hours or days at a time, with a 220mb
>>>> cache_mem. Now, however the memory usage grows to 900mb in ~40
>>>> minutes, and typically reaches 1.5gb in ~4 hours. We have a ulimit in
>>>> place to kill it once it hits 1.5gb, but prior to putting that in
>>>> place it typically reached 2gb.
>>>>
>>>>>
>>>>> If you can provide more details I will be happy to try and test it.
>>>>>
>>>>> Thanks,
>>>>> Eliezer
>>>>
>>>>
>>>> If there's any other information you think may be useful, feel free to
>>>> ask.
>>
>>
Received on Mon Dec 23 2013 - 20:22:24 MST