tis 2009-09-15 klockan 10:58 -0400 skrev Sachin Malave:
But if we use external library like openmp then i dont think we need
to make much changes in actual codes. But that is only possible if we
find blocks in existing codes which could be executed by different
threads...
Unfortunately the
tis 2009-09-15 klockan 23:59 +1200 skrev Amos Jeffries:
+if ( (tmp_sock = socket(PF_INET6, SOCK_STREAM, IPPROTO_TCP))
= 0
+setsockopt(tmp_sock, SOL_IP, IP_TRANSPARENT, (char
*)tos, sizeof(int)) == 0
+bind(tmp_sock, (struct sockaddr*)tmp_ip6, sizeof(struct
mån 2009-09-14 klockan 14:53 +1200 skrev Amos Jeffries:
Proposed fix: Have src/ssl_support.h include et/com_err.h in an extern
C wrapping before OpenSSL includes...
So... does it actually require the extern-C ?
com_err.h yes. It's a C header with no built-in C++ shielding, which
means any
mån 2009-09-14 klockan 22:43 +1200 skrev Amos Jeffries:
on epoll ( select ) implementations in squid. It is found that epoll
is polling all the descriptors processing them one by one. There is
an important FD used by http port which is always busy, but has to
wait for other descriptors
mån 2009-09-14 klockan 06:02 -0700 skrev Paul Khadra:
Dear All,
2009/09/14 15:30:32| clientNatLookup: NAT lookup failed: ioctl(SIOCGNATL)
The above message is appearing every minute in the cache.log
Which means Squid and your NAT kernel do not talk properly to each
other, making the original
mån 2009-09-14 klockan 02:04 -0600 skrev Alex Rousskov:
Apparently, we are packing StoreEntry.swap_file_sz into the stored
entry header before we compute its value. This happens because to
compute swap_file_sz, we need to know swap_hdr_sz, the size of the
stored entry header. To compute
mån 2009-09-14 klockan 11:27 +0200 skrev Guido Serassio:
Done. Sorry for the delay, but I'm very busy (as usual ...)
I have placed the preprocessed source of two failing files
(IpIntercept.cc and QosConfig.cc) here:
http://www.acmeconsulting.it/libip.zip
Odd... std::string::npos is
seems to build fine for me on CentOS 5.2.. but maybe something in the
test builds are triggering it..
A full test is running..
sön 2009-09-13 klockan 13:14 +0100 skrev Markus Moeller:
Could this be a g++ problem ? error_messge is defined as external in
et/com_err.h and I see that -lcom-err
sön 2009-09-13 klockan 21:30 +0200 skrev Henrik Nordstrom:
seems to build fine for me on CentOS 5.2.. but maybe something in the
test builds are triggering it..
A full test is running..
Found.. it fails when used in combination with --enable-ssl. This due to
OpenSSL (or actually krb5.h I
lör 2009-09-12 klockan 23:36 +1200 skrev Amos Jeffries:
Updating the checklist today I again wonder if we can repeat the step
from 2.7 and enable HTTP/1.1 on requests sent to servers
The default in 2.7 is 1.0 still. There is an option to enable 1.1, or
actually three.. (http11 cache_peer
tor 2009-09-10 klockan 14:53 +1200 skrev Amos Jeffries:
On Thu, 10 Sep 2009 01:03:48 +0100, Markus Moeller
hua...@moeller.plus.com wrote:
How about this then ? Please replace the squid_kerb_auth directory with
the
No luck with the error_message function changes yet.
ons 2009-09-09 klockan 13:13 +1200 skrev Amos Jeffries:
How exactly are you accessing the HEAD code to generate these patches?
None of them has applied cleanly by the time they got to me.
Same here, seems whitespace gets munged somewhere...
Regards
Henrik
ons 2009-09-09 klockan 06:27 +0100 skrev Markus Moeller:
Sorry about that. I use rsync://squid-cache.org/source/squid-3 as noted here
http://www.squid-cache.org/Versions/v3/HEAD/ .Maybe I forgot to update.
Hmm.. seems that got stuck at 5 nov 2008 should be fixed now.
Regards
Henrik
ons 2009-09-09 klockan 19:49 +0100 skrev Markus Moeller:
I still see 5 Nov 2008.
Right, only the branches got fixed, still failed to update the HEAD
versions...
Now that's fixed as well.
Regards
Henrik
ons 2009-09-09 klockan 23:45 +1200 skrev Amos Jeffries:
Amos Jeffries wrote:
Update of cvs.devel.squid-cache.org:/cvsroot/squid/website
In directory fdv4jf1.ch3.sourceforge.com:/tmp/cvs-serv18699
Modified Files:
projects.html
Log Message:
Remove some dropped/completed
-I/home/henrik/build/fedora/squid/devel/squid-3.1.0.13'
The -R/usr/lib64 and -I/usr/include options are both unasked for.
Regards
Henrik
tis 2009-09-08 klockan 01:01 +0100 skrev Markus Moeller:
How about the attached ?
Markus
Henrik Nordstrom hen...@henriknordstrom.net
tis 2009-09-08 klockan 19:05 +0100 skrev Markus Moeller:
I don't remember exactly why I added it. I think it was because I originally
had it as a standalone configure with the option to have libs and includes
in exec_prefix/lib and exec_prefix/include where exec_prefix is the squid
install
mån 2009-09-07 klockan 12:18 +1200 skrev Amos Jeffries:
Tricky. That would place the hash at the wrong end of the file (last) where
its most likely to be overlooked. Particularly on the longer config files.
Doesn't matter if it's overlooked.
There are also some distros (notably Gentoo and
mån 2009-09-07 klockan 11:52 +1200 skrev Amos Jeffries:
This seems to be related:
http://www.squid-cache.org/bugs/show_bug.cgi?id=2719
I do't think so. Works for me.
Regards
Henrik
mån 2009-09-07 klockan 19:51 +1200 skrev Amos Jeffries:
Henrik Nordstrom wrote:
mån 2009-09-07 klockan 12:18 +1200 skrev Amos Jeffries:
Tricky. That would place the hash at the wrong end of the file (last) where
its most likely to be overlooked. Particularly on the longer config files
mån 2009-09-07 klockan 19:51 +1200 skrev Amos Jeffries:
But here is another idea. Have uninstall compare with the source
directory and not the target. Would probably be best. Will screw up if
someone tries make clean before make uninstall but that's their
problem. And if that's a problem
mån 2009-09-07 klockan 00:16 +1200 skrev Amos Jeffries:
ERROR: files left after uninstall:
./etc/mime.conf
./etc/squid.conf
make[1]: *** [distuninstallcheck] Error 1
make[1]: Leaving directory
getnameinfo.h util.h)
Marksu
Henrik Nordstrom hen...@henriknordstrom.net wrote in message
news:1252072098.571.16.ca...@henriknordstrom.net...
lör 2009-09-05 klockan 01:33 +1200 skrev Amos Jeffries:
Markus,
these changes won't help the Fedora build with Squid-3.1 frozen
mån 2009-09-07 klockan 07:08 +1000 skrev Robert Collins:
On Sun, 2009-09-06 at 22:52 +0200, Henrik Nordstrom wrote:
Yes.
With no dependencies to guide it make reorders as it sees fit for the
day, especially if running parallel jobs.
We should be fine if we just list
mån 2009-09-07 klockan 07:29 +1000 skrev Robert Collins:
On Sun, 2009-09-06 at 23:25 +0200, Henrik Nordstrom wrote:
Not sure I am comfortable with adding dependencies to automakes own
targets..
and no, we are not fine with just that. See the rest of previous
response.. To repeat
fre 2009-09-04 klockan 17:57 +1200 skrev Amos Jeffries:
-.po.lang:
+%.lang:
Hmm... that SHOULD have worked with any POSIX make.
% rules do not unfortunately.
Did the automake on Centos pick up the suffixes to generate the
matching .SUFFIXES: rule?
Regards
Henrik
lör 2009-09-05 klockan 01:33 +1200 skrev Amos Jeffries:
Markus,
these changes won't help the Fedora build with Squid-3.1 frozen. That
will require a minimal change of probably just the configure.in.
Squid-3.1 has now been packaged for Fedora 12, but so far without
squid_kerb_auth due to
Maybe that it's C++ while the others are C?
Hmm.. from Clang homepage
The Clang Static Analyzer consists of both a source code
analysis framework and a standalone tool that finds bugs in C
and Objective-C programs..
C++ is not C or Objective-C, so it's not too strange it
ons 2009-09-02 klockan 21:11 +0200 skrev Guido Serassio:
Hi,
OK, but what next ?
Can you produce a preprocessed source? That's the output of just the
preprocessor, not actual compilarion, comparable to gcc -E option.
Easier to identify what the compiler actually saw that way...
Regards
ons 2009-09-02 klockan 12:42 -0700 skrev Guy Bashkansky:
Now I see the Expires header having a value in the past, which may
confuse clients and caches further down the chain.
Scenario: origin returns max-age=900 (15 min) and refresh_pattern
overrides expire to 24 hours, what do the headers to
ons 2009-09-02 klockan 13:42 -0700 skrev Guy Bashkansky:
I'm using a customized version of Squid 2.4 STABLE6. But nothing
seems to be customized in refresh.c, except for my own recent swap of
age and expires checks (as recommended).
Ouch.. that's a very very old relese. Upgrading highly
bb: approve
+1. Sane, and correct, and also matches what Squid-2 is doing.
Hmm.. wonder how that got lost in Squid-3. The change in squid-2 is very
old (3 may 2000). Investigating. Right, got lost in refactoring
(revision 5998). The comment from when this was fixed in Squid-2 is even
left just
tis 2009-09-01 klockan 12:49 +0200 skrev n...@squid-cache.org:
bzr: ERROR: Invalid http response for
http://www.squid-cache.org/bzr/squid3/.bzr/repository/indices/ee58751feae77ae5e40d734321628f2c.tix:
Expected a boundary
(squid/3.1.0.13~fischeri20090822:651DED8031766650F2BFAF4D8497827E)
Needs quoting:
+KRB5INCS=`$krb5confpath --cflags krb5 2/dev/null`
+KRB5LIBS=`$krb5confpath --libs krb5 2/dev/null`
(seen twice, Solaris generic)
Would also be nice if you could update squid_kerb_auth/configure with
this simplified kerberos configure dance. The
mån 2009-08-31 klockan 22:02 +1200 skrev Amos Jeffries:
NP: not quite sure about the incrementation. Write appears to be called
once per log entry but this is not tested under enough conditions to be
certain yet. The old lineEnd call from Squid-2 is not present in Squid-3.
Well, for
ons 2009-08-26 klockan 18:17 -0700 skrev Guy Bashkansky:
If indeed refresh_pattern only extends expiration, I would like to
develop a feature that enforces an exact time-to-live (per URL) in my
local branch of Squid code.
See the refreshStaleness() function. Should be sufficient to move the
hnordstrom has voted tweak.
Status is now: Conditionally approved
Comment:
Looks fine, but could use some minor cleanup like the #if nesting in
include/helper_debug.h.
Also the use of helper_debug() vs debug() seem a little inconsistent.
(the #if added to
ons 2009-08-26 klockan 02:32 +1200 skrev Amos Jeffries:
I kind of agree. This is a step towards cleaning up the permutations a
bit. The old ones are not really pass-thru. As you note they all did
some synthesizing which can under some circumstances is mutually
exclusive with real
ons 2009-08-26 klockan 07:19 +1000 skrev Robert Collins:
Even the smallest embedded devices are pretty huge these days, so the
extra libs don't concern me.
And trivial to explicitly disable if needed.
Regards
Henrik
The code dealing with our special case auth header
forwarding/synthesising had grown quite hairy and partially duplicated
from all additions.
# Bazaar merge directive format 2 (Bazaar 0.90)
# revision_id: hen...@henriknordstrom.net-20090826115054-\
# 5xsrhlscmmkjgr84
# target_branch:
ons 2009-08-26 klockan 00:01 +1200 skrev Amos Jeffries:
Hmm, I'm thinking there should not be too much functional change to make
this fix Bug-7 in a slightly better way than squid-2 did.
Oh, but there is...
all this change does is to allow older objects to get back into memory.
It does not
From what I can tell the difference between the PASSTHRU and PASS is
only that PASSTHRU do not add any injected credentials from
external_acl, right?
Imho there is no need for more than two of these options.
PASS - WWW+Proxy authentication passed along as-is if present.
external_acl auth added
t
Theres a few bits of polish I know still needed:
* some of the DiskIO dependencies I could not find clearly remain
outside the scan block.
* according to Robert the Windows Overlapping IO needs to be broken
out into its own module. That is another cleanup sub-project.
And the build
tor 2009-08-20 klockan 09:00 +1000 skrev Robert Collins:
I think we should have an infrastructure product in bugzilla, for
tracking list/server/buildfarm etc issues.
My vote is for an other product.
Regards
Henrik
hnordstrom has voted approve.
Status is now: Approved
Comment:
Looks fine.
But the patch highlighted another minor issue.. we should get the hier
strings autogenerated from the enum. Having lists like that in the
source is quite errorprone.
For details, see:
hnordstrom has voted comment.
Status is now: Semi-approved
Comment:
Is this still relevant?
For details, see:
http://bundlebuggy.aaronbentley.com/project/squid/request/%3C4A003331.1050406%40treenet.co.nz%3E
Project: Squid
hnordstrom has voted tweak.
Status is now: Semi-approved
Comment:
Better to turn it the other way around. Add a new header to be used as
substitute for Cache-Control by the reverse proxy.
For details, see:
fre 2009-08-21 klockan 13:55 +1200 skrev Amos Jeffries:
Aye, spent some time looking for such in the code before discovering we
did not already do it.
Anyone have good ideas how to do that?
We already have mk-string-arrays.awk converting selected enums from
enum.h to string arrays. Should
ons 2009-08-19 klockan 16:15 +1200 skrev Amos Jeffries:
Henrik, this is the new memory-promotion patch dying.
Does not look like the patch, but I will look into it regardless.
My first reaction is Why is that assert there?. May well be the case
that it should not be asserted.
Regards
Henrik
ons 2009-08-19 klockan 19:20 +0200 skrev Henrik Nordstrom:
ons 2009-08-19 klockan 16:15 +1200 skrev Amos Jeffries:
Henrik, this is the new memory-promotion patch dying.
Does not look like the patch, but I will look into it regardless.
In fact it was... For some reason I tried to optimize
mån 2009-08-17 klockan 14:08 +1200 skrev Amos Jeffries:
´
For this usage 127.* is not a bogon at all.
Yet 0.0.0.0 in it's place would be completely insane despite any
trickery the TCP stack might do to cope.
Is it? It either goes to 127.0.0.1 or nowhere..
Regards
Henrik
sön 2009-08-16 klockan 18:20 +1200 skrev Amos Jeffries:
Henrik Nordstrom wrote:
revno: 9907
committer: Henrik Nordstrom hen...@henriknordstrom.net
branch nick: trunk
timestamp: Sat 2009-08-15 14:56:39 +0200
message:
Add
sön 2009-08-16 klockan 19:17 +1200 skrev Amos Jeffries:
Aha. Just connect() then? not really bind() or listen()?
Correct. Bind to 0.0.0.0 is any address.
I'm thinking that aliasing has already been done before Squid gets such
packets at the 'other end'. So that we only see the real
sön 2009-08-16 klockan 21:42 +0800 skrev Adrian Chadd:
There's more build failure messages on squid-dev then actual
development discussion.
Yes, as there is a release notes change some revisions back blocking the
builds...
Perhaps the build failure email should start spamming the person who
sön 2009-08-16 klockan 03:31 -0600 skrev Serassio Guido:
Windows port: Update mswin_check_ad_group to version 2.0
Does this mean you are ready for release now?
Regards
Henrik
sön 2009-08-16 klockan 23:09 +0200 skrev Guido Serassio:
I think that the new helper should be stable enough, so you can start with
the STABLE 7 release process.
Done. Next nightly is most likely the code for 2.7.STABLE7. But remains
to update releasenotes, changelog etc...
Regards
Henrik
sön 2009-08-16 klockan 11:19 +0200 skrev Stephen R. van den Berg:
Well, let's put it this way: Linux says (I checked) that the RSS
(Resident Set Size) is around 18MB. This seems a bit wasteful for
something that is just sitting idle with a minimal memory cache and no
disk cache.
For what
mån 2009-08-17 klockan 08:20 +1000 skrev Robert Collins:
I thought I'd just gauge interest in having 2.HEAD and 2.CURRENT tested
in the buildfarm. For all that most development is focused on 3, there
are still commits being done to 2.x, and most of the hard work in the
buildfarm is setup -
lör 2009-08-15 klockan 00:42 +0200 skrev Stephen R. van den Berg:
I recently started using squid again after a very long time.
Welcome back!
Checking the memory usage I see:
Memory accounted for:
Total accounted: 1850 KB 6%
memPool accounted: 1850 KB 6%
lör 2009-08-15 klockan 09:45 -0400 skrev Matt W. Benjamin:
I wonder if the build farm messages could go to a special list?
I think squid-dev is quite fit for the purpose, but as already discussed
the volume of these notification will be reduced shortly to just one
message and not one per host
lör 2009-08-15 klockan 01:28 +1200 skrev Amos Jeffries:
Amos Jeffries wrote:
Port of the hier_code ACL from 2.7.
This goes a little further than a straight port by adding Henriks
suggestion of a numeric-based array of codes.
Instead of a string-wise search on every use it's a speedy
tor 2009-08-13 klockan 13:40 -0600 skrev Alex Rousskov:
Henrik, please commit if you do not hear any objections. I am looking
forward to benchmarking this important change.
Done.
Regards
Henrik
fre 2009-08-14 klockan 16:10 +1200 skrev Amos Jeffries:
Do we want to start planning for when the memory cache settings are all
options on the one base tag?
ie. memory_cache size options
where options ::= mode replacement_policy max-object-size
min-object-size ... etc.
That's a
Just a thought, but to avoid spamming with failed reports when a
platform-independent test is forgotten to be updated / fails, would it
be possible to configure Hudson to use one build host as master (i.e. a
reasonably new Linux one), and defer trying the other if the master is
reported as failed?
lör 2009-08-15 klockan 09:34 +1000 skrev Robert Collins:
This would cause builds on other platforms to wait for the master to
suceed.
Which is what is desired I think. No need to waste energy or noise on
builds which will most likely fail.
Any other drawbacks?
Regards
Henrik
lör 2009-08-15 klockan 13:26 +1200 skrev Amos Jeffries:
Now closed it, but I wonder, does this close or simply reduce the impact
of bug #7?
Reduces it sligthly as a sideeffect, but I would not account for it..
Am I right in thinking the in-memory copy gets demoted at some later
point back
squid.conf option for tuning when to keep objects in memory.
Sponsored by: The Measurement Factory
# Bazaar merge directive format 2 (Bazaar 0.90)
# revision_id: hen...@henriknordstrom.net-20090813102833-\
# 32fj8b4jw78a2dph
# target_branch: ../memcache
# testament_sha1:
fre 2009-08-14 klockan 11:23 +1200 skrev Amos Jeffries:
cache_mem is where the fastest HITS/sec come from. Trashing it will
impact the very upper req/sec of Squid. As will forcing everything to
disk first. Then again with things going disk-memory (how soon Henrik?)
it may not be noticeable
This looks wrong to me.. Netfilter is not TPROXY and does not require
libcap.
The libcap requirement test for TPROXY is further down I think, or at
least it looked so from the rest of the diff..
ons 2009-08-12 klockan 21:38 +1200 skrev Amos Jeffries:
+dnl Netfilter TPROXY depends on libcap
sön 2009-08-09 klockan 12:38 +1200 skrev Amos Jeffries:
Worst-case;
adding slow dstrdns which always converts whats given to IP and then
converts that to rDNS may be a possibility.
I'd rather see an option to dstdomain to disable rDNS lookups if
undesired. It by default does a reverse
fre 2009-08-07 klockan 16:27 +1200 skrev Amos Jeffries:
Huh? I think you misunderstand what these alias lists do.
They are the seed maps for make install to generate such symlinks from.
Oh, I thougt we loaded them.. oh well.. probably we should but that's
another question.
The upgrade from
Why are we even having all these aliases? In most cases it should be
sufficient with the base language alone.
Language tags may contain all sorts of subtags, with the subtags
changing relatively frequently. Having each and every possible
combination listed isn't very realistic.
See
ons 2009-08-05 klockan 21:54 +1200 skrev Amos Jeffries:
Right now it's to improve the performance of the negotiation, each
detected but failed language requires disk IO.
Right. it is reading the error page from disk each time. Forgot that
little detail..
When sub-tag negotiation is reliably
ons 2009-08-05 klockan 01:01 +0800 skrev Adrian Chadd:
G'day,
I just noticed in src/HttpReply.c that the vary expire option
(Config.onoff.vary_ignore_expire) is checked if the reply has HDR_VARY
set but it does not check if HDR_X_ACCELERATOR_VARY is set.
Everywhere else in the code
Vidarebefordrat meddelande
Från: Robert Mattson r.matt...@latrobe.edu.au
Till: l...@lists.mozilla.org, i...@squid-cache.org,
us...@httpd.apache.org, www-t...@w3.org
Ämne: HTTP-MPLEX
Datum: Mon, 3 Aug 2009 17:29:32 +1000
Dear Firefox, Squid, Apache and W3 communities,
tis 2009-08-04 klockan 14:19 -0700 skrev Mark Nottingham:
I think we need to get him in touch with the HTTP-over-SCTP folks...
Agreed.
Regards
Henrik
tis 2009-08-04 klockan 23:32 +0200 skrev Henrik Nordstrom:
tis 2009-08-04 klockan 14:19 -0700 skrev Mark Nottingham:
I think we need to get him in touch with the HTTP-over-SCTP folks...
Agreed.
Done.
Regards
Henrik
mån 2009-08-03 klockan 07:32 +1000 skrev Robert Collins:
Are youo coordinating with Andrew Bartlett on getting
overlapping/concurrent stuff in the winbind helpers?
Not at the moment. Not even on the track of adding support for this to
Squid yet. Just stumbled on the deferred stuff while
mån 2009-08-03 klockan 11:41 +1200 skrev Amos Jeffries:
Oh yeah, on that front.
I've contemplated and think it would be very useful to publish a little bit
of exemplar logic for third-party helpers to detect whether or not Squid is
sending concurrent requests. That way the third-parties are
mån 2009-08-03 klockan 22:01 +1200 skrev Amos Jeffries:
The idea I was going with was just a simple test for the numeric, so
that whether or not the admin has configured it does not kill the helper
and shortly Squid. We still get those complains occasionally.
Unfortunately doesn't work that
mån 2009-08-03 klockan 12:11 +0100 skrev Markus Moeller:
The attached patch is against the latest cvs and brings squid 2 in line with
squid 3.
Applied.
Regards
Henrik
mån 2009-08-03 klockan 13:05 +0100 skrev Markus Moeller:
As long as squid_kerb_auth has a bigger buffer then defined in
auth_negotiate.c(c) it should be fine. The theortical maximum is 64k if I
remember right for data in an uathentication header or ?
Current 3.0 limit is 64KB for whole
sön 2009-08-02 klockan 20:49 +1200 skrev Amos Jeffries:
I've been holding off on that one Henrik.
That's fine for 3.1, but trunk should push this a little imho.
Until the loadable modules issues Debian/Ubuntu have with libtool 2.2.6a
are resolved one way or the other it's basically fatal
Looking at the stateful helpers code I noticed we still have the
deferred concept, even if it's not used at all.
As part of preparing the code for concurrent helper protocol for
NTLM/Negotiate I propose dropping the deferred support, and possibly the
helper resets as well as there isn't really
mån 2009-08-03 klockan 07:11 +1000 skrev Robert Collins:
There's still going to be the concept of 'tcp connection X needs auth
helper Y to send its next step on'; this is essentially deferred
requests.
No, that's the RESERVED state..
The DEFERRED state were used by the challenge-reuse code
lör 2009-08-01 klockan 16:41 +0100 skrev Markus Moeller:
In some setups the upstream proxy requires a secue authentication method
(Negotiate, NTLM). The attached patches (2.7 and 3.0) allow this with
Negotiate.
Imported to Squid-2 with the following cosmetic modifications
* Your
tor 2009-07-30 klockan 10:26 +1000 skrev Robert Collins:
In corporate networking, TLS MITM is a 'feature': company signed
certificates are used to sign the TLS connection to the corporate
firewall, and the firewall validate the SSL connection to the outside
world. I haven't personally used
tor 2009-07-30 klockan 08:21 + skrev Ian Hickson:
1) CONNECT and HTTP Upgrade are optional, and independent. One may be used
or
the other. They may be both tried in any order.
That doesn't sound specific enough to get interoperable behaviour. It
seems like we'd want to define
tis 2009-07-28 klockan 20:42 +1200 skrev Amos Jeffries:
IIRC you knew something about the reserved/deferred state requirements
for NTLM helpers.
A bit.. having rewritten that code about 1.5 times.
I'm looking at bug 2648, where the helpers are idle in RESERVED pinned
state and causing
fre 2009-07-24 klockan 14:24 -0600 skrev Alex Rousskov:
-if (hdr_len Config.maxReplyHeaderSize || (hdr_len = 0
(size_t)buf-contentSize() Config.maxReplyHeaderSize)) {
+if (hdr_len = Config.maxReplyHeaderSize || (hdr_len = 0
(size_t)buf-contentSize()
ons 2009-07-22 klockan 11:08 +0200 skrev Guido Serassio:
I'm totally aware of all the un-merged code in the nt branch.
Excellent.
Currently there are mainly two different un-merged works in the nt branch:
1. The release 2.0 of mswin_check_ad_group external acl helper on
which I'm
Hi Amos,
it seems the function which extracts info from the releasenotes don't
always do the right thing..
http://www.squid-cache.org/Doc/config/half_closed_clients/
Regards
Henrik
tis 2009-07-21 klockan 03:51 -0700 skrev EliG:
I am making my firts steps in Squid... and trying to see the content of the
Html Response.
In a future I will try to modify that content.
I think the eCAP interface may be more suited for your goals, but you
may do it within the Squid code as
Hi Guido,
I just activated the SourceForge branch viewer again when realizing it's
been off for 1.5 years, and noticed a bit of code has accumulated up in
the nt branch of Squid-2. I guess some (most) of that should be
committed back to the main repository?
From a quick and possibly incomplete
sön 2009-07-19 klockan 16:06 +0800 skrev Adrian Chadd:
I've fixed a potentially risky situation in Lusca relating to the
initialisation of the storeIOState cbdata type. Each storedir has a
different idea of how the allocation should be free()'ed.
Risky in what sense?
All your patch does is
lör 2009-07-18 klockan 20:02 +1000 skrev Robert Collins:
So the patch from gentoo - s/CXX_HOST/CXX_FOR_BUILD/ - and its nearly
right. The only problem is determining the build machine cxx. sadly
autoconf and automake don't have good built in foo for this (at least
according to their manuals).
fre 2009-07-17 klockan 16:32 +1200 skrev Amos Jeffries:
3.1.5.1
Connect to the server given by /host/ on port 80 and ask it to
upgrade from HTTP to WebSockets.
The client must send only the following lines:
GET /thepath/ HTTP/1.1
Host:
fre 2009-07-17 klockan 10:00 + skrev Ian Hickson:
HTTP servers aren't the only concern, of course; we want the handshake to
be as secure as possible against any other protocol that may exist on any
server that may be deployed. We don't know what's out there (especially in
intranets),
tor 2009-07-16 klockan 11:13 +1000 skrev Robert Collins:
Note that with NTLM this is realistic - there's no need for multiple
helpers as a single helper can serve many concurrent requests.
No longer the case as we no longer do challenge reuses and the
concurrent protocol is not implemented for
tor 2009-07-16 klockan 14:47 +1200 skrev Amos Jeffries:
And the sad reality in squid-3 with _one_ helper:
squid-*--1-helper---* winbindd [state1, state2]
\---crash: all 2 of 1 helpers all RS pending state.
You can't run NTLM with just one helper, not until both the
101 - 200 of 2137 matches
Mail list logo