On Fri, 2014-03-28 at 10:48 +, Andrew Beverley wrote:
> Minor updates to the qos_flows documentation. Clearer instructions
> based on recent squid-users mailing list questions.
Apologies, please find attached updated patch. I have made clearer the
requirement to use CONNMARK not MARK.
Hi guys,
Please find attached patch for minor updates to the qos_flows
documentation, based on recent squid-users questions.
Andy
Minor updates to the qos_flows documentation. Clearer instructions
based on recent squid-users mailing list questions.
=== modified file 'src/cf.data.pre'
--- src/cf
On Mon, 2012-01-23 at 11:42 +1300, Amos Jeffries wrote:
> The current website layout for Versions/ and Download/ is a little
> confusing and I would like to combine the two into a simpler form.
...
> Comments? Other ideas?
>
+1 from me. I have hunted around several times for the link to the
re
On Mon, 2011-11-07 at 11:59 +1300, Amos Jeffries wrote:
> Well, in light of the facts that new helpers are only being added to
> 3.3 now
That's fair enough. I've attached my (fairly raw) helper to this email
anyway, just for the list archives, in case anyone else has use for it.
> and that li
On Mon, 2011-11-07 at 13:55 -0700, Alex Rousskov wrote:
> FWIW, the usual approach (in a helper or elsewhere) is to reread the
> configuration file only when the file modification time stamp changes
> and/or a HUP signal is received, and not on every request.
Thanks, I hadn't thought of that. Now
On Sun, 2011-11-06 at 14:17 +1300, Amos Jeffries wrote:
> On 6/11/2011 1:39 p.m., Andrew Beverley wrote:
> > Hi,
> >
> > I am using the ACL feature whereby the parameters can be read from a
> > file. For example:
> >
> > acl session_day time "/var/www/an
Hi,
I am using the ACL feature whereby the parameters can be read from a
file. For example:
acl session_day time "/var/www/announce_days.txt"
Understandably, the file only appears to be read when the configuration
file is parsed, rather than each time the ACL is checked. However, I
need it to be
Hi,
I'm getting occasionally Squid segfaults (from the worker I think) with
version 3.2.0.13. During busy periods these can be several times an
hour, at other times it could be once or twice a day.
What is the best way of tracing the problem? Should I increase the
logging verbosity and see if I c
On Tue, 2011-11-01 at 15:22 +1300, Amos Jeffries wrote:
> On Mon, 31 Oct 2011 23:33:47 +0000, Andrew Beverley wrote:
> > Hi,
> >
> > The attached patch adds a mask option to the qos_flows miss
> > configuration value. The reason for this is to allow the preserved
> &g
On Tue, 2011-11-01 at 12:12 +1300, Amos Jeffries wrote:
> On Mon, 31 Oct 2011 22:19:12 +0000, Andrew Beverley wrote:
> > On Mon, 2011-10-31 at 15:42 -0600, Alex Rousskov wrote:
> >> On 10/31/2011 03:03 PM, Andrew Beverley wrote:
> >>
> >> > Having thought
Hi,
The attached patch adds a mask option to the qos_flows miss
configuration value. The reason for this is to allow the preserved
mark/TOS value from the server to be altered slightly rather than
overwritten completely.
Example usage. The following will preserve the netfilter mark, but will
ensu
On Mon, 2011-10-31 at 15:42 -0600, Alex Rousskov wrote:
> On 10/31/2011 03:03 PM, Andrew Beverley wrote:
>
> > Having thought about this further, I think what I was trying to achieve
> > was getting the mark every time a packet was received.
>
>
> > I have tried
On Mon, 2011-10-17 at 10:03 +0200, Kinkie wrote:
> On Mon, Oct 17, 2011 at 5:30 AM, Amos Jeffries wrote:
> > On 17/10/11 11:23, Andrew Beverley wrote:
> >>
> >> Hi,
> >>
> >> I've been using the qos_flows feature for preserving a ne
On Mon, 2011-10-17 at 10:34 -0600, Alex Rousskov wrote:
> On 10/16/2011 04:23 PM, Andrew Beverley wrote:
>
> > I've been using the qos_flows feature for preserving a netfilter mark,
> > but have run into some problems.
> >
> > Currently, the netfilter mar
Sorry for the delay, stand by for further emails on this subject as well
as the comments below...
On Mon, 2011-10-17 at 16:30 +1300, Amos Jeffries wrote:
> On 17/10/11 11:23, Andrew Beverley wrote:
> > Hi,
> >
> > I've been using the qos_flows feature for preserving a ne
On Sun, 2011-10-09 at 20:06 +0100, Andrew Beverley wrote:
> On Sat, 2011-10-08 at 21:18 +0200, Henrik Nordström wrote:
> > fre 2011-10-07 klockan 19:18 +0100 skrev Andrew Beverley:
> >
> > > I admit that I am rushing to submit this as I go away for the weekend,
> > &
Hi,
I've been using the qos_flows feature for preserving a netfilter mark,
but have run into some problems.
Currently, the netfilter mark for the connection is obtained in
forward.cc, during the stages of opening a connection to the remote
server. The problem with this is that the connection mark
On Sat, 2011-10-08 at 21:18 +0200, Henrik Nordström wrote:
> fre 2011-10-07 klockan 19:18 +0100 skrev Andrew Beverley:
>
> > I admit that I am rushing to submit this as I go away for the weekend,
> > so please let me know if it's not up to scratch! Works For Me (TM)
>
Further to discussions, please find attached a patch for the session
helper to:
- Remove support for Berkeley DB 1.85
- Add support for the current Berkeley DB (db.h)
- Add support for a DB environment (fixes synchronisation between
multiple processes)
- Fix the active mode bug previously submitte
On Thu, 2011-10-06 at 13:23 +1300, Amos Jeffries wrote:
> > Oh, I forgot to say: the DB_ENV functionality requires a *directory*
> > rather than a file to be specified for the database (it creates
> > several
> > files). So, although I don't see a problem with this in principle, it
> > does break
On Wed, 2011-10-05 at 23:41 +0100, Andrew Beverley wrote:
> I have tried using the same functions on the modern API: this didn't fix
> the problem (probably because it is just running in compatibility mode).
>
> I then tried the modern API, but using the added features (DB_ENV).
On Thu, 2011-10-06 at 12:07 +1300, Amos Jeffries wrote:
> I think 1.85 is so ancient now its not really needed for old-OS
> support. The problem is more likely to be adding #if to support db.h and
> 4.2 or such for RHEL.
Do you mean Berkeley DB 4.2? If so, from the little I have read, I think
> > FWIW, the ssl_crtd daemon which stores generated SSL certificates
> > on
> > disk does [lock and] reopen the database every time it needs to read
> > it.
> > This is not efficient, but avoids conflicts and stale info.
> >
> > If you do not like that simple but inefficient approach,
It wa
On Tue, 2011-10-04 at 18:59 +0100, Andrew Beverley wrote:
> However, I'm now having problems with multiple instances of the session
> helper writing to the same database. I thought I had fixed this with the
> ->sync option, but it appears not. If I open multiple instances of
>
On Sat, 2011-10-01 at 15:51 +1300, Amos Jeffries wrote:
[...]
> When the LOGIN/LOGOUT ACLs are tested they perform their action on the
> session state. The ACLs leading up to them have to be crafted to avoid
> testing them at all unless you want LOGIN/LOGOUT to happen on that
> request
[
On Thu, 2011-09-29 at 13:49 +0200, Henrik Nordström wrote:
> tis 2011-09-27 klockan 07:32 +0100 skrev Andrew Beverley:
>
> > I'd like to find a way around this. The best way that I can think of is
> > to add an option to the session helper, to specify a URL that must
So, I've been using the session helper for a few days now to display a
splash page, and on the whole it's working well.
However, one of the problems that I have experienced is that there is
often something in the background on a user's computer that retrieves
something from the web and forces a re
On Wed, 2011-09-21 at 07:54 -0600, Alex Rousskov wrote:
> On 09/20/2011 11:29 PM, Amos Jeffries wrote:
> >>>
> >>> In file included from ../src/ssl/support.h:38,
> >>> from ssl/ErrorDetailManager.h:4,
> >>> from errorpage.cc:42:
> >>> ../src/ssl/gadgets.h:39: err
On Sat, 2011-09-24 at 15:39 +0200, Henrik Nordström wrote:
> mån 2011-09-19 klockan 10:49 +1200 skrev Amos Jeffries:
>
> > The session helper in Squid-3 is concurrent. The user_key is the opaque
> > channel-ID. (Probably should be renamed to match the protocol
> > documentation).
> > http:/
On Wed, 2011-09-21 at 11:01 +1200, Amos Jeffries wrote:
> >> > Done. Also, the following page should be updated:
> >> >
> >> > http://wiki.squid-cache.org/ConfigExamples/Portal/Splash
> >> >
> >> > I'm happy to do it myself, if you can give me wiki edit rights?
> >>
>
> Enabled.
>
I've updated
On Thu, 2011-09-22 at 11:10 +1200, Amos Jeffries wrote:
> > However, it still leaves the question: what is the best way to log
> > errors from the helper. At the moment, even with the patch applied, a
> > user will still get "The helpers are crashing too rapidly, need help" if
> > they don't specif
On Wed, 2011-09-21 at 11:01 +1200, Amos Jeffries wrote:
> On Tue, 20 Sep 2011 22:26:20 +0100, Andrew Beverley wrote:
> > On Tue, 2011-09-20 at 12:09 +1200, Amos Jeffries wrote:
> >> > Do you mean the -d option to the Squid binary? If so, this doesn't
> >> >
On Tue, 2011-09-20 at 12:09 +1200, Amos Jeffries wrote:
> > Do you mean the -d option to the Squid binary? If so, this doesn't
> > seem
> > to make any difference; it just prints all the log messages to the
> > display as well as the log file.
>
> -d parameter of the helper binary. stderr is pip
On Mon, 2011-09-19 at 10:49 +1200, Amos Jeffries wrote:
> The session helper in Squid-3 is concurrent.
Ah, okay.
> The user_key is the opaque
> channel-ID. (Probably should be renamed to match the protocol
> documentation).
> http://wiki.squid-cache.org/Features/AddonHelpers#Access_Contro
Hi,
I have run into a problem using the session helper (ext_session_acl). In
its current format, the session helper expects 2 parameters as a
minimum. However, using the example at
http://wiki.squid-cache.org/ConfigExamples/Portal/Splash only one is
passed (the IP address).
The second parameter e
On Sun, 2011-06-05 at 18:12 +0200, Henrik Nordström wrote:
> sön 2011-06-05 klockan 17:04 +0100 skrev Andrew Beverley:
> > Firstly please tell me if I should be posting this to the users mailing
> > list...
> >
> > I am trying to set up a splash page using the ext_sess
Firstly please tell me if I should be posting this to the users mailing
list...
I am trying to set up a splash page using the ext_session_acl helper.
However, I get the error message "FATAL: The ext_session_acl helpers are
crashing too rapidly, need help!" in the log.
I tried initially using V3.1
On Sat, 2011-05-07 at 17:15 +1200, Amos Jeffries wrote:
> Two months ago we set a goal/checkpoint of early May to have 3.2 on the
> last legs toward release. This is it. Are we on track? No.
>
Would it be possible to get the attached patch applied for the release
please? This makes the compilati
> > Taking a closer look at the yes/no/auto logics and teh particular reason
> > for changing it I think that is a bug in the SQUID_DEFINE_BOOL. I'm
> > proposing a different simpler change in other discussion thread.
> >
>
> That bit is now has a simpler fix in trunk. You can remove the changes
> >
> > Personally I am quite fine with requiring pkg-config as a build
> > requirement for automtic detection of libcap, openssl, openldap and a a
> > couple more. My only requirement is that a minimal build should be
> > possible even without pkg-config.
> >
> > pkg-config is often available even
On Mon, 2011-01-10 at 22:37 +1300, Amos Jeffries wrote:
> On 10/01/11 19:58, Andrew Beverley wrote:
> > Hi all,
> >
> > I was recently caught out by my own patch when compiling Squid :-)
> > I compiled with netfilter marking enabled, but couldn't work out why
>
On Mon, 2011-01-10 at 22:26 +1300, Amos Jeffries wrote:
> On 10/01/11 19:58, Andrew Beverley wrote:
> > Hi all,
> >
> > I was recently caught out by my own patch when compiling Squid :-)
> > I compiled with netfilter marking enabled, but couldn't work out why
>
Hi all,
I was recently caught out by my own patch when compiling Squid :-)
I compiled with netfilter marking enabled, but couldn't work out why
packets weren't being marked. It was only after turning on detailed
logging that I realised it was because Squid had been compiled without
libcap.
Theref
> http://wiki.squid-cache.org/Squid3CodingGuidelines already covers both C++
> and automake policies. I've added a menu to easily navigate the page and
> this under autoconf guidelines.
>
Hmmm, my fault entirely for not looking properly, but I didn't realise
that page existed. Can I suggest a li
On Wed, 2010-10-06 at 10:17 +0200, Kinkie wrote:
> > Hmm, we at least want to MSG_NOTICE for both cases, with preferrably a hard
> > error if its explicitly stated.
> >
> > This is where the yes/no/auto comes in handy to switch the type of fail
> > message.
>
> Hi guys,
> handling of the --with-n
On Mon, 2010-10-04 at 18:12 +1300, Amos Jeffries wrote:
> On 20/09/10 00:41, Andrew Beverley wrote:
> >>>>> I've moved it next to the headers check. I have also removed the error
> >>>>> message that was generated if they don't exist. However, th
On Sun, 2010-09-19 at 04:24 +1200, Amos Jeffries wrote:
> On 19/09/10 00:47, Andrew Beverley wrote:
> > On Sat, 2010-09-18 at 20:34 +1200, Amos Jeffries wrote:
> >> On 18/09/10 09:18, Andrew Beverley wrote:
> >>> Hi,
> >>>
> >>> Please find a
On Thu, 2010-09-16 at 16:23 -0600, Alex Rousskov wrote:
> On 09/15/2010 12:12 AM, Andrew Beverley wrote:
> > On Wed, 2010-09-15 at 02:06 +, Amos Jeffries wrote:
> >> On Tue, 14 Sep 2010 23:55:20 +0100, Andrew Beverley
> >> wrote:
> >>
On Wed, 2010-09-15 at 19:44 +1200, Amos Jeffries wrote:
> On 15/09/10 18:12, Andrew Beverley wrote:
> > On Wed, 2010-09-15 at 02:06 +, Amos Jeffries wrote:
> >> On Tue, 14 Sep 2010 23:55:20 +0100, Andrew Beverley
> >> wrote:
> >>
On Wed, 2010-09-15 at 02:06 +, Amos Jeffries wrote:
> On Tue, 14 Sep 2010 23:55:20 +0100, Andrew Beverley
> wrote:
> >> * Config.accessList.outgoingTos, Config.accessList.clientsideTos,
> >> Config.accessList.outgoingNfmark, Config.accessList.clientsideNfmark
> can
> * Config.accessList.outgoingTos, Config.accessList.clientsideTos,
> Config.accessList.outgoingNfmark, Config.accessList.clientsideNfmark can
> become members of the Qos scope Config object. All the parsing /free stuff
> can be moved there too with some #define parse_...() etc for the legacy
> p
> src/Parsing.*:
> * Please palce the strtou*() functions to a file under lib/ with an .h in
> include/. At this stage they get linked in globally through lib/libmisc.la.
Done, but as they both return bool I've had to add stdbool.h to the
headers that the file includes and also add a check for s
On Fri, 2010-09-10 at 16:46 +1200, Amos Jeffries wrote:
> On 10/09/10 07:08, Andrew Beverley wrote:
>
> >
> >> * ARG_WITH if-yes clause then becomes:
> >>case "$withval" in
> >> yes|no) with_netfilter_conntrack=$withval ;;
> >>
On Mon, 2010-09-06 at 02:53 +, Amos Jeffries wrote:
> On Sun, 05 Sep 2010 21:59:34 +0100, Andrew Beverley
> wrote:
> > Please find attached the latest version of the patch to add Netfilter
> > marking support to Squid.
> >
> > All the previous comments have n
> * I find the terminology inconsistent and confusing: outgoing,
> clientside, upstream. No wonder you have to explain the difference
> twice. Unless these are all standard RFC-like terms, please use
> something consistent like fromClient, toClient, fromServer, toServer.
> Others may suggest a
Hi Scott,
> Anyway, joining this list because I am scoping a project which involves
> implementing a captive portal to manage Internet access for unknown
> users (the general public) for a "free wifi" hotspot.
>
> I realise I'm not the first person to set up a captive portal for a free
> wifi
On Mon, 2010-09-06 at 23:06 +0100, Andrew Beverley wrote:
> > >
> > >> * I find the terminology inconsistent and confusing: outgoing,
> > >> clientside, upstream. No wonder you have to explain the difference
> > >> twice. Unless these are all standar
> >
> >> * I find the terminology inconsistent and confusing: outgoing,
> >> clientside, upstream. No wonder you have to explain the difference
> >> twice. Unless these are all standard RFC-like terms, please use
> >> something consistent like fromClient, toClient, fromServer, toServer.
> >> Others
Thanks for the (very) prompt response (I'm impressed). Replies to other
comments to follow; in the meantime...
> * I find the terminology inconsistent and confusing: outgoing,
> clientside, upstream. No wonder you have to explain the difference
> twice. Unless these are all standard RFC-like ter
> > The above configure concept would tie in with removing the --enable-qos
> > option altogether. There's no reason for the QOS code not to be included
> > that I can see (it has no dependencies, apart from the optional upstream
> > kernel patch), and with this patch and the isTosActive(), it's e
On Fri, 2010-08-13 at 18:19 -0600, Alex Rousskov wrote:
> On 08/11/2010 03:25 PM, Andrew Beverley wrote:
>
> > I've moved these, as well as most of the other QOS functions, into
> > Ip::Qos. I have also removed the QosConfig namespace, as it didn't seem
> >
> >
> > * Do you need an L suffix for large unsigned constants like 0x?
> > Please investigate. I do not know the answer, but I recall seeing such
> > suffixes elsewhere:
> > http://www.google.com/search?q=0x+vs+0xL
>
> I thought that indicated "long" type to be used. Wh
My latest revision of the patch for netfilter marking will follow soon.
Before I post it, I will reply to comments on the previous version.
> >> >
> >> > Question number 2: what is stubQosConfig.cc? Does that also need
> >> > updating for this patch?
> >> >
> >>
> >> stub* are cut down set of
Quick question please:
Should variables be named in the form outgoingNetfilterMark or
outgoing_netfilter_mark, or does it not matter? There appears to be a
variety of formats in use in the existing code.
Thanks,
Andy
> * My understanding is that class data members and public class methods
> should be documented in the header. Others should be documented in the
> .cc files. You may want to double check this rule with Amos before
> moving comments though.
>
> * Many Qos data members are not documented, inclu
On Fri, 2010-08-20 at 12:44 -0600, Alex Rousskov wrote:
> On 08/20/2010 11:06 AM, Andrew Beverley wrote:
> > On Fri, 2010-08-13 at 18:19 -0600, Alex Rousskov wrote:
> >> On 08/11/2010 03:25 PM, Andrew Beverley wrote:
> >>
> >>> I've moved these, as w
On Fri, 2010-08-13 at 18:19 -0600, Alex Rousskov wrote:
> On 08/11/2010 03:25 PM, Andrew Beverley wrote:
>
> > I've moved these, as well as most of the other QOS functions, into
> > Ip::Qos. I have also removed the QosConfig namespace, as it didn't seem
> >
> Seems the netfilter guys found a major problem with strtoul().
> Thankfully the same fix should work for us as well.
>
>
> Luciano Coelho wrote:
>
> >
> > Not easily. I found that there is a bug in strtoul (and strtoull for
> > that matter) that causes the long to overflow if there are
> > Firstly, I'm slightly confused as to whether I am seeing the
> > autoconf-refactor work in my copy of trunk (I have run bzr update). I
> > didn't think I'd seen it go in, but delving into configure.in I think it
> > might be in there... what should I look for to check whether it's
> > definitel
Andrew,
> I have compile squid for the Mac OSX 10.6.4 and is working with no issue
> with reguards too squid.conf setting , I run Privoxy --> Squid -->
> tor=internet but when I look at the cache.log it is filling up with errors.
> This seem too have happened on on the openbsd some time ago and ne
Hi,
Firstly, I'm slightly confused as to whether I am seeing the
autoconf-refactor work in my copy of trunk (I have run bzr update). I
didn't think I'd seen it go in, but delving into configure.in I think it
might be in there... what should I look for to check whether it's
definitely what I'm look
> >> stub* are cut down set of all non-inline Ip::QosConfig methods and any
> >> globals defined in QosConfig.h. Changes to the API need to be mirrored
> >> there. The functions inside usually call fatal() to alert a wrong
> >> linkage clearly during testing. In this particular case the parse
>
Updated patch attached; comments below.
> > If we can move to strtoul, I would like to change 'tos' to char
> > throughout. Currently it is possible to set it to invalid values in
> > squid.conf, which then cause problems with dumpConfigLine.
> >
> > Question number 2: what is stubQosConfig.cc? D
Thanks for the prompt response. Updated patch attached to my previous
email.
> > Please find attached the first version of the netfilter mark patch. I've
> > not yet tested it extensively, but would welcome some initial feedback
> > or comments.
>
> The mess around local port can be cleaned up (s
On Mon, 2010-08-02 at 12:03 -0600, Alex Rousskov wrote:
> On 08/01/2010 05:47 PM, Andrew Beverley wrote:
> > Please find attached the first version of the netfilter mark patch. I've
> > not yet tested it extensively, but would welcome some initial feedback
> > or commen
Please find attached the first version of the netfilter mark patch. I've
not yet tested it extensively, but would welcome some initial feedback
or comments.
My comments are:
- The existing TOS patch cannot be disabled at runtime. As such, this
mark patch cannot be either. Would it be preferable t
> I'm not sure its fully worth doing this.
>
> * the "transparent" options are all due for a naming upgrade or
> removal in the next major release anyway.
Okay, I'm happy to leave as is. However, I would still suggest a review
of the naming in the upgrade (see below).
>
> * linux-netfilter i
I'd like to propose the attached patch, to rename the build option
--enable-linux-netfilter to --enable-nf-transparent. This is for 2
reasons:
1. It is consistent with the remainder of the transparent proxy options
(ifpw-transparent, ipf-transparent, pf-transparent).
2. It causes less confusion w
Please find attached very minor patch for configure.in
# Bazaar merge directive format 2 (Bazaar 0.90)
# revision_id: a...@andybev.com-20100731150544-p9g5vwcakwomkug9
# target_branch: file:///home/andrew/squid-repo/trunk/
# testament_sha1: acdd91fe4f43f7f08ff31f224b583c5310f8398c
# timestamp: 2010
I'm currently editing configure.in for my proposed QOS mark patch. From
a previous list message Amos suggested the following:
>> --enable/disable-linux-netfilter will also be involved in the logics.
If set to "no" then it override disables this netfilter feature. <<
I started to implement this, b
> So, do you have a clear use-case we can add to the wiki and commit
> message?
I propose extending the current QualityOfService feature as follows. The
existing http://wiki.squid-cache.org/Features/QualityOfService page
should read:
* Allows you to set a TOS/Diffserv value to mark local and p
On Wed, 2010-07-14 at 23:35 +0100, Andrew Beverley wrote:
> > > In order to obtain mark information from the existing connection (using
> > > libnetfilter_conntrack), I need to know the local and remote port
> > > number, and the local and remote IP address of each con
> > In order to obtain mark information from the existing connection (using
> > libnetfilter_conntrack), I need to know the local and remote port
> > number, and the local and remote IP address of each connection. Most of
> > this information is in 'class fde', but not all of it. Is it available
>
As per my previous posts, I'm working on a patch to implement the ZPH
features of Squid, but with packet marking (use-case to follow once I'm
sure I can achieve what I want to).
In order to obtain mark information from the existing connection (using
libnetfilter_conntrack), I need to know the loca
> > So, is the best way of implementing this to do the same as transparent
> > proxying, and check whether the (proposed) marking option is enabled in
> > squid.conf when executing restoreCapabilities? If the user has asked for
> > packets to be marked, then CAP_NET_ADMIN will be retained. The mark
> > I have done some initial scoping, but have discovered that in order to
> > mark a packet using setsockopt(), the process needs to be run as root.
>
> Are you sure it needs root and not just a suitable capability flag? From
> what I can tel CAP_NET_ADMIN is sufficient.
You're right, it only ne
> > 1. Because the marking process needs to be run as root, can this only be
> > achieved by putting the mark function within the squid process that
> > originally starts up, and stipulate that this has to be run as root?
>
> Consider a dedicated helper like the diskd helper - send it a fd using
>
I am considering writing a patch for Squid so that it maintains a
packet's netfilter mark value if not fetched from the cache. This would
be similar to the QOS functionality, in that there would also be an
option to set the mark on a packet that is fetched from the cache.
I have done some initial
88 matches
Mail list logo
