server v2.example.com
and so on
Thank you.
-Paras
On Fri, May 2, 2008 at 9:58 PM, Henrik Nordstrom
[EMAIL PROTECTED] wrote:
On ons, 2008-04-30 at 23:51 +1200, Amos Jeffries wrote:
size_t is named to describe its use. It's unsigned to record the length
of things.
IIRC
I'd suggest you extend the key to allow it to include an IP to get the
network and individual stats as well..
but it's fine as it is..
On mån, 2008-06-16 at 22:49 +0800, Adrian Chadd wrote:
Patch against Squid-2.HEAD:
http://www.creative.net.au/diffs/20080716-delay_snmp-1.diff
On fre, 2008-06-13 at 11:29 +0800, Adrian Chadd wrote:
Ok, so where would you draw the line between client/server side stuff and
storage stuff?
With or without major cleanup?
Is it enough to be able to store multiple variants of the same object on
some arbitrary keyset; and then place the
On tor, 2008-06-12 at 09:56 +0800, Adrian Chadd wrote:
Yes, but there's more to Vary processing than this bug - eg, the
unable to establish vary_id for XXX, possible Vary storage/retrieval
improvements, etc.
unable to.. is one of the effects of this bug.. Squid hitting an
internal state which
On ons, 2008-06-11 at 01:18 +0800, Adrian Chadd wrote:
Henrik,
Whats incomplete about the Variant invalidation support in PURGE/CLR in
Squid-2.HEAD?
There are many common corner cases where the existing vary_id isn't
picked up properly. The worst being if an object goes from not having
Vary
On tis, 2008-06-10 at 12:40 +0300, Mikko Kettunen wrote:
How could I configure squid to buffer whole POST-request before sending
it to peer cache or originserver?
Good question. It isn't really designed to do this. POST request may be
almost unbounded in size, and some streaming over HTTP
On ons, 2008-06-11 at 17:48 +0800, Adrian Chadd wrote:
Subject says it all - open, read and unlink are async, but write and close
aren't.
Why is at least write not async?
Simply because when testing on Linux they very rarely block, with the fs
buffers absorbing them fully as long as you don't
On ons, 2008-06-11 at 20:01 +0800, Adrian Chadd wrote:
What I need to do is get this Sun box actually running Solaris and try to
get dtrace to tell me what time is spent in the main thread blocking on
disk writes.
On Solaris you defenitely want async writes and closes, as Solaris UFS
by
On tis, 2008-06-10 at 15:45 +1000, Mark Nottingham wrote:
Because peerGetSomeParent and then peerGetAllParents are both called
in the NEVER_DIRECT case (e.g., accelerators), the list of peers to
use will contain a duplicate; whatever peerGetSomeParent returns will
be in the list twice,
On lör, 2008-06-07 at 10:43 +0800, Adrian Chadd wrote:
I think some of their stuff was backed out of Squid-2.7 before
release.
The Vary invalidation patch was backed out from 2.7 as it's incomplete
and broke things.
But this X-Vary-Options patch never got committed. Thread stops after
your
On fre, 2008-06-06 at 15:38 +1000, Mark Nottingham wrote:
Logging the client-side request done point is an option, but it's of
limited value, because most requests don't have a body, and as
mentioned, the server can start sending the response before the
request is done. Not against it,
Interested.
Ideally one would like to get rid of the in-stream transport features of
HTTP, moving message delimitation fully to the transport (both positive
and negative delimitation). Hopefully without introducing too much
additional overhead ontop op SCTP.
Regards
Henrik
On ons, 2008-06-04
()
On 04/06/2008, at 11:59 PM, Henrik Nordstrom wrote:
On tis, 2008-06-03 at 14:17 +1000, Mark Nottingham wrote:
I'm getting the following with 2.7HEAD-20080602:
2008/06/02 21:08:17| assertion failed: cbdata.c:239: c-y ==
CBDATA_COOKIE(p)
A stacktrace would help a lot
The documentation on how to use the different logfile modules seems to
be missing in 2.7.
Regards
Henrik
On lör, 2008-05-31 at 12:36 -0700, Pranav Desai wrote:
But I feel that I am missing something here. So, I would really
appreciate any suggestions or comments you may have.
To me it souns like you are very much on track. An option to http_port
specifying the protocol to listen on is the right
On tis, 2008-05-27 at 09:42 +0800, Adrian Chadd wrote:
Hi,
Here's a small diff (thanks to benno!) which implements a simple fast
ACL lookup for the maximum request size. This allows Squid administrators
to restrict POST/upload sizes based on a variety of available request
information.
On lör, 2008-05-24 at 03:23 +0200, Henrik Nordstrom wrote:
the nightly snapshot of trunk is failing with
make: don't know how to make testheaders.c. Stop
Fixed. test-tools/Makefile.am had testheaders.sh listed as a built
binary instead of as a distributed file.
Also fixed other problems
On fre, 2008-05-23 at 15:47 +0200, Cezary Rzewuski wrote:
I've looked into squid's code and I've got an idea how to do this. The
best place to scan downloaded site seems to be storeSwapOutFileClosed
function in store_swapout.cc file. After closing file clamav could scan
this file and log if
On fre, 2008-05-16 at 23:17 +0200, Sven Köhler wrote:
As far as i can tell, the server is Limecast 2.0.0. When the squid-Proxy
is being used, the mp3-stream is terribly broken. Without squid
inbetween and the server and my client (mplayer, audacious and the like)
the stream works fine.
On lör, 2008-05-17 at 08:25 +0200, Sven Köhler wrote:
Yes, that was what i was thinking of when saying handle gracefully.
Are these patches available for squid 3.0 stable too?
Not yet.
Regards
Henrik
On lör, 2008-05-17 at 20:15 +0200, Sven Köhler wrote:
Is there some option, to turn off HTTP 0.9 transformation to HTTP 1.0?
This would already help A LOT in my case.
There is none, but shouldn't be too hard to add I guess, making it
revert more or less to the old behaviour.
When such
On lör, 2008-05-17 at 20:32 +0200, Sven Köhler wrote:
So actually, the mp3stream is found and played by my mp3-players. Yet,
mp3 stream has hearable errors in it. Is it possible, that there is
data-loss or that data gets modified somehow?
Is there some transformation of the data taking
On sön, 2008-05-18 at 02:49 +0800, Adrian Chadd wrote:
On Sat, May 17, 2008, Henrik Nordstrom wrote:
`-Wsign-compare'
I have this flag as a candidate warning flag in my build script, but not
activated as there is a bit of work to prune out the false alerts..
Where about
On lör, 2008-05-17 at 21:42 +0200, Sven Köhler wrote:
But at the moment, i'm completely irritated.
When i run a wget http://streamserver; first, and then play the file
with mplayer, everything sounds OK.
But when i run mplayer http://streamserver;, then the mp3 errors occur,
and the
On lör, 2008-05-17 at 22:29 +0200, Sven Köhler wrote:
Actually, i'm close to reproducing the problem.
I've setup a simple server at port 80 (xinetd+shellscript) which replies
something like ICY 200 OK\r\n\r\n[mp3stream] to any connect on port
80. And now i think that i can show data
On tis, 2008-05-13 at 20:56 +1200, Amos Jeffries wrote:
This patch reverts the behaviour to what was reportedy available in 2.5.
Adding a file named cachemgr.conf.default next to the cachemgr.conf
2.5 never had cachemgr.conf.
The Debian version did, but that was a distro local
On mån, 2008-05-12 at 01:01 +1200, Amos Jeffries wrote:
revno: 8988
committer: Amos Jeffries [EMAIL PROTECTED]
branch nick: trunk
timestamp: Mon 2008-05-12 01:01:38 +1200
message:
Author: Thomas-Martin Seck [EMAIL PROTECTED]
On lör, 2008-05-10 at 20:41 +1200, Amos Jeffries wrote:
This Feature was scheduled for 2.8.
Having upgraded the squid-3 patch I think it is small enough to go into 3.1
now.
+1
Trivial feature addition to make the settings make sense in an area
which has been very confusing..
Regards
Henrik
On tor, 2008-05-08 at 08:45 -0600, Alex Rousskov wrote:
Since the code is already in 2.6, the patch can probably go back into
3.0 if Henrik has time to apply the changes he wants and the code is
still not disruptive after that.
It is not in 2.6, but the feature will appear in 2.7.
Regards
On tor, 2008-05-08 at 07:43 -0700, Ben Drees wrote:
My SourceForge username is bendrees.
Congratulations. You now have access to the public developer CVS
repository, and may host any Squid development of your choice there.
Regards
Henrik
On fre, 2008-05-09 at 10:23 +1000, Mark Nottingham wrote:
Yes, assuming that a) I'm running squid-3, and b) I'm willing to go
through the pain of running an ICAP server, writing the plugin, etc.
etc. etc.
Neither is true.
Running an ICAP and writing the needed rules for it is not much
I don't think the zph patch belongs in 3.0.STABLE. It's a new feature
with several new directives, not a bugfix, and additionally I'd like a
bit of cleanup before release.
Saw that it got merged to 3.0 some day ago... and I vote that it gets
backed out from 3.0. But should stay in trunk.
Regards
On mån, 2008-05-05 at 18:26 +0200, Guido Serassio wrote:
It's since few months that icons are missing in the cvsweb on
www.squid.cache.org:
http://www.squid-cache.org/cgi-bin/cvsweb.cgi/
Hunting for lost icons... ah, it uses the icons shipped with Apache. I
guess that alias got lost when
On sön, 2008-05-04 at 09:48 -0700, Christian Seifert wrote:
Now, however, Id like to go a step further and cache independent of
query param value. Again, a request to www.foo.com/test?q=1 is cached
and now upon a second request with the same param, but different
value, such as
On sön, 2008-05-04 at 18:10 -0600, Cron Daemon wrote:
Tree is up to date at revision 8975.
60586: old priority 0, new priority 10
automake :
autoconfg:
libtool :
Bootstrapping
[...]
x - extracting anthony-xbm.gif (binary)
x - extracting anthony-xpm.gif (binary)
make: don't know how
On fre, 2008-05-02 at 21:58 +1200, Amos Jeffries wrote:
I'll add to that: some testers seem to be unable to pull revisions later
than 8963 out of trunk.
The anonymous http mirror of the repository had stalled and the mail
with the error went into /dev/null somewhere. Fixed this the other day
ons 2008-04-30 klockan 15:58 +0800 skrev Adrian Chadd:
I figured out how to extract sensible callgraph oprofile reporting out of a
64 bit
install.
Notes are here:
http://code.google.com/p/cacheboy/wiki/CodeProfiling
Basically - it seems the omit-frame-pointer optimisation GCC does
4 clause BSD license is not GPL compatible due to the advertising
requirements in '3''.
But you should be able to find this file with a 3 clause BSD license.
On sön, 2008-04-27 at 09:33 +, Guido Serassio wrote:
Update of cvs.devel.squid-cache.org:/cvsroot/squid/squid
Modified Files:
On sön, 2008-04-27 at 20:20 +0200, Guido Serassio wrote:
Now I'm very confused.
Don't be. BSD removed clause 3 from their license some many years ago
after some discussion with GPL folks and others, which resulted in the 3
clause BSD license normally seen today. You just happened to find a copy
On sön, 2008-04-27 at 16:50 +0200, Guido Serassio wrote:
Hi,
In the way to merge into 2.HEAD all the Windows port stuff, one of
the remains is the addition of reconfigure control to cachemgr. This
functionality was available in the Windows port since 2.5, and for
Windows admins is very
On sön, 2008-04-27 at 21:36 +0200, Guido Serassio wrote:
But we have a problem for lib/radix.c and
lib/inet_ntoa.c, they are also licensed with the old BSD license ...
Should be fine to delete clause 3 from these as well from what I can
tell.
Regards
Henrik
2.6.STABLE20 has now ben rerolled and released (not yet signed, will do
that tomorrow). This is intended to be the last Squid-2.6 release before
2.7.
Regards
Henrik
there is something wrong with the bzr repository currently,
investigation pending. Some operations result in a error from bzr about
an invalid packet '\r'.
additionally somehow the SQUID_3_0 branch got lost for a while today,
acting as if it was equal to trunk. Not 100% sure how that happened,
tor 2008-04-24 klockan 22:54 +0200 skrev Henrik Nordstrom:
there is something wrong with the bzr repository currently,
investigation pending. Some operations result in a error from bzr about
an invalid packet '\r'.
The corrupted has been restored, and the non-bzr event which causes
ons 2008-04-23 klockan 07:57 +0200 skrev Maliling List, Westerhold,
Axel:
'Will not work' comes down to will not work through ICAP when then
responding server incorrectly uses Chunked enconding on HTTP 1.0 from squid.
Squid-3 trunk has implemented chunked encoding decoding, and should get
tis 2008-04-22 klockan 09:47 +0800 skrev Adrian Chadd:
Store the hit information in the MemObject; include it in swap metadata.
It's about misses so no need to swap it out..
Storing it in the mem object is acceptable. It really should go into the
reply, but you then have to worry abour ICAP
tis 2008-04-22 klockan 10:40 +0200 skrev Guido Serassio:
This is the version included into the MinGW runtime, I think that it
should be fine:
http://cygwin.com/cgi-bin/cvsweb.cgi/src/winsup/mingw/mingwex/dirent.c?cvsroot=src
Public domain is fine, even if it's debated if authors really can
tis 2008-04-22 klockan 15:48 +0200 skrev Guido Serassio:
Wow, trying to fix a Squid copyright issue, I have found a problem
into another open source project, I'm very lucky :-(
Yes.. it's fun to look into licensing issues..
The following is the FreeBSD version, it's more simpler to
tis 2008-04-22 klockan 19:35 +0800 skrev Adrian Chadd:
yeah, thats pretty invalid. :) I wonder if your method for returning
RSS is valid.
RSS is Resident Set Size, the number of in-core memory pages currently
occupied by the process. Not the process size, or even amount of used
memory for this
tis 2008-04-22 klockan 18:01 +0200 skrev Guido Serassio:
As you know I'm working to the native Windows port of the support for
the resolv.conf domain directive, and working on it I have also found
some little memory leak in the Windows code, so I was hoping to
include the changes into
ons 2008-04-23 klockan 00:58 +0800 skrev Adrian Chadd:
In any case, Guido's response of its the number of pages! explained the
base discrepancy. The unfortunate issue here is that pagesize can be
variable
and one would need to find the system/process' pagesize (and hope its
constant!)
to
mån 2008-04-21 klockan 10:04 +0200 skrev Guido Serassio:
But before, I like to focus on Squid 2 before the 2.7 release, hoping
that in the meantime the bzr people will fix the problem
As good plan as any. Porting 2.7 should be very straight forward from
2.6.
I like to run some Windows
mån 2008-04-21 klockan 12:57 +0200 skrev Guido Serassio:
What happens when both search and domain keywords are specified into
resolv.conf ?
The last one is used.
It seems to me that the last parsed overwrites the domain search list.
Yes. As it does in the glibc resolver..
Regards
Henrik
mån 2008-04-21 klockan 16:27 +0200 skrev Guido Serassio:
But there is some standard defined about ?
Doubtful. resolv.conf is not specified in SUS, so the closest to
standard for resolv.conf is the bind implementation I think.
This could be a GLIBC bug
Could be. Not sure domain is meant
mån 2008-04-21 klockan 21:35 +0200 skrev Axel Westerhold:
--- src/http.cc 2008-04-01 13:54:38.0 +0200
+++ src/http.cc 2008-04-21 19:11:47.0 +0200
@@ -1200,7 +1200,7 @@
void
HttpStateData::maybeReadVirginBody()
{
-int read_sz = replyBodySpace(readBuf-spaceSize());
+
lör 2008-04-19 klockan 21:01 -0600 skrev Alex Rousskov:
I agree with many of your specific suggestions (and will discuss them
separately), but I want to understand what problem you are proposing to
solve first. For example:
- Fix a few new bugs introduced by async calls.
- Fix certain comm
lör 2008-04-19 klockan 22:14 -0600 skrev Alex Rousskov:
FWIW, I have always wanted to get rid of COMM_ERR_CLOSING in Squid3 :-).
A close callback is sufficient to let the FD user know that there will
be no more callbacks for that FD.
Yes.
With some effort, we may be able to
pragmatically
sön 2008-04-20 klockan 22:19 +0300 skrev Christos Tsantilas:
The important files are the CommCalls.* AsyncCalls.* and ICAP/AsyncJob.*
But I believe that you are understand it enough.
I think I understand most of the code now, but not entirely 100% sure on
the design and intended use model
sön 2008-04-20 klockan 22:14 -0600 skrev Alex Rousskov:
The current suggestions are:
.dox (in docs/Programming Guide or in src/, not clear) -- Amos.
.dox (in src/) -- Kinkie.
.h (or perhaps .cc) -- Henrik.
To clarify:
Longer API documents, .dox file in docs/, or maybe src/ next to the .cc
sön 2008-04-20 klockan 22:56 -0600 skrev Alex Rousskov:
- I hope we can avoid similar problems with comm_cancel_read because the
two or three places using that call can be adjusted to cancel the read
callback directly. We have already talked about that elsewhere.
It's not only
fre 2008-04-18 klockan 08:25 -0600 skrev Alex Rousskov:
The changes needed on the client-side are pretty much unrelated to async
calls. Async calls just make the design problems more obvious and help
in solving them. Removing async calls from comm will fix a few bugs, but
will be a step
Tor 2008-04-17 klockan 21:19 +0530 skrev Paras Fadte:
With respect to C++ , is a static member of a class like a global
variable?
Yes, a class scoped (not object) global variable.
Does OOP in particular consider use of a static/global
variable as a bad programming practice?
Most do..
tis 2008-04-15 klockan 20:34 +0200 skrev Guido Serassio:
Except for trunk, stuck at 20 March, it seems that all daily snapshot
are no more build since the 6 April.
All snapshots seem to work now..
Regards
Henrik
tis 2008-04-15 klockan 11:56 +0800 skrev Adrian Chadd:
The profiling abilities of current-generation tools is pretty damned
sweet. I'd really suggest investigating them over the internal profiling
stuff nowdays.
+1, and something I raised when XPROF_STATS was added..
Regards
Henrik
tis 2008-04-15 klockan 20:30 +0200 skrev Guido Serassio:
I cannot waste my very limited time trying to fix the development
tools that I should use . :-(
My proposal if you find that you have time to work on Squid-3, ignoring
the tools problem:
Create an NT branch in the devel CVS
sön 2008-04-13 klockan 03:13 +1200 skrev Amos Jeffries:
I always use: grep -R blah src/*
So either is fine, and the latter does feel more organized.
Same here, or variants thereof.
Regards
Henrik
mån 2008-04-14 klockan 07:15 +1000 skrev Robert Collins:
Anyone up for doing this...
In file included from /home/robertc/source/squid/trunk/lib/hash.c:64:
/home/robertc/source/squid/trunk/include/profiling.h:59:2: error:
#warning Unsupported CPU. Define function get_tick(). Disabling
mån 2008-04-14 klockan 10:24 +0200 skrev Guido Serassio:
In theory on x64 systems we could use TSC like on x86, but:
http://developer.amd.com/assets/TSC_Dual-Core_Utility.pdf
http://h2.www2.hp.com/bizsupport/TechSupport/Document.jsp?objectID=c01075682
Current generation of Intel CPU is
tis 2008-04-15 klockan 01:55 +1200 skrev Amos Jeffries:
As for the auto-notificatin of these problems. Whomever is running the
daily crontab might want to pass the stderr output to noc.
The snapshots is currently running on my account, but should probably be
moved to a service account...
sön 2008-04-13 klockan 12:28 +1200 skrev Amos Jeffries:
Aha. I think I've got it going.
http://treenet.co.nz/projects/squid/branches/tproxy4
The bzr repository looks fine.
Haven't reviewed the changes as such yet.
Regards
Henrik
fre 2008-04-11 klockan 01:00 -0600 skrev Alex Rousskov:
Receiving or handling a call is not an instantaneous event though (and
never was). Consider this pseudo-code:
Yes. And at Call::dial calling foo.handleThissCall() I consider the call
made. The fact that foo.handleThisCall() hasn't
fre 2008-04-11 klockan 15:42 +1200 skrev Amos Jeffries:
Round 2 of polished Transparency/Interception upgrade of Squid-3.
This patch and bundle follows the same model as previously submitted but
also includes fallout fixes from TPROXY testing.
IPFW which was also modified in the polish
fre 2008-04-11 klockan 10:52 -0600 skrev Alex Rousskov:
A comm_cancel_my_read wrapper will help mitigate risk (but not eliminate
it). Note that once the comm_cancel_my_read is in place, the only
mistake the programmer can make is to call Call::cancel instead of
calling comm_cancel_my_read.
tor 2008-04-10 klockan 01:23 +0300 skrev Tsantilas Christos:
The submitted pach didn't work for two reasons. In fact this
cancellation of comm scheduled calls doesn't work at all.
Are you getting the same assertion with your patch, or a different
one? (I am not geeting this assertion
tis 2008-04-08 klockan 18:58 -0600 skrev Alex Rousskov:
In general, the caller should forget about the call after making it
(i.e., after scheduling it). We can make comm different, but I am not
sure it is a good idea from the clarity/simplicity point of view.
Which kind of rules out using
ons 2008-04-09 klockan 17:36 +0800 skrev Adrian Chadd:
The ipt_bind call from forward.c should go into comm_openex. Kind of
pointless with the new flag otherwise.. You said you had done this, but
not in the -3 patch...
That was done but I wasn't diff'ing comm.c
Don't forget to remove it
mån 2008-04-07 klockan 23:41 -0600 skrev Alex Rousskov:
Can you post cache.log lines with full debugging enabled? I would be
happy to help with the queue review, but it would speed things up if I
do not have to go through the complicated steps required to reproduce
the problem.
The cache.log
tis 2008-04-08 klockan 23:50 +1200 skrev Amos Jeffries:
If its not easy and its not simple it means more bugs. I don't think 'c'
is a good choice here without a full ground-up recode. Nay on that from
me on general principles.
True. I view 'c' as more of a last resort if the other turn out
tis 2008-04-08 klockan 13:23 +1200 skrev Amos Jeffries:
Is there a security problem with them being set?
IMO its just a cleanup otherwise (already voted those not to go back).
It's not even a source change, just a bzr attributes cleanup. But the
current situation may confuse some unexperienced
tis 2008-04-08 klockan 17:59 +1200 skrev Amos Jeffries:
Currently:
fde::flags::transparent == 'intercept/non-intercept'
fde::flags::tproxy == real-transparent/non-transparent
(new) COMM_TRANSPARENT == real-transparent
Their use is currently good for what they do. A small re-naming
tis 2008-04-08 klockan 14:28 +0800 skrev Adrian Chadd:
Take 2: includes initial modularisation (untested; I'll test it at home
this weekend when I get my tproxy box back online) and configure magic
(with placeholders for tproxy4/freebsd.)
tis 2008-04-08 klockan 13:23 +1200 skrev Amos Jeffries:
+1. I'm just waiting on you all to agree that its tested enough. If you
want to do the merge yourself Henrik, I'm okay with that.
Doesn't matter for me who merges the backport. It's a trivial bzr merge
with the backport already done..
tis 2008-04-08 klockan 22:17 +0800 skrev Adrian Chadd:
* make sure upstream/peer-forwarded requests aren't thrown to the tproxy
code.
I read the current code as don't do that; its the behaviour I'd like to
maintain. We can always add the functionality later.
Yes, it is not something
tis 2008-04-08 klockan 13:57 +0100 skrev Bradley Kite:
Any hints/tips that you can recomend on a low budget? I've looked at
Web Polygraph, and in a lab environment this seems excellent, but I
think it requires more hardware than I have available to me.
There is a number of other tools simpler
tis 2008-04-08 klockan 08:57 -0600 skrev Alex Rousskov:
IIRC, one-bucket-at-a-time was too slow in environments that cared more
about hits than optimizing miss performance. Folks wanted a knob to
control the trade-off.
Ok. We should change that to be buckets based, and default to 1.
The
tis 2008-04-08 klockan 07:47 -0600 skrev Alex Rousskov:
I believe there is a better option that was added to handle specifically
to handle these kind of problems:
d) Cancel the call when it is no longer expected.
That's also an option, not at all far from 'a'.
This is already supported by
ons 2008-04-09 klockan 00:19 +0800 skrev Adrian Chadd:
On Tue, Apr 08, 2008, Adrian Chadd wrote:
Take 3:
http://www.creative.net.au/diffs/20080408-tproxy-fix-2.diff
using -3 instead of -2 looks better... but incomplete.
You need to add the new ipt_* files to the EXTRA_squid_SOURCES
tis 2008-04-08 klockan 20:46 +0300 skrev Tsantilas Christos:
AsyncCalls are refcounted so believed that with ccb-callback = NULL
the AsyncCall which the ccb-callback point released. But yes if it is
refcounted by an other object (for example in AsyncCalls queues,)
it is not released and
in commSetTimeout there is the following fragment which looks very odd
to me
if (callback != NULL) {
typedef CommTimeoutCbParams Params;
Params params = GetCommParamsParams(callback);
params.fd = fd;
F-timeoutHandler = callback;
}
tis 2008-04-08 klockan 12:55 -0600 skrev Alex Rousskov:
BTW, if somebody commits that patch, please polish the reason phrase in
the cancel() call to something more specific.
The submitted pach didn't work for two reasons. In fact this
cancellation of comm scheduled calls doesn't work at all.
tis 2008-04-08 klockan 17:55 -0600 skrev Alex Rousskov:
is worth the bugs and complexity we are _removing_. Also, the future SMP
support may operate within the same nothing-is-immediate problem space.
In other words, let's get used to it! :-)
Note: For SMP to work out you want as few hops
tis 2008-04-08 klockan 18:04 -0600 skrev Alex Rousskov:
Yes, it is setting the fd member of the call[back].
Yes, I realized that after sending the message.
The syntax is very ugly, but I do not have a better proposal at the
moment. I'll guess I get used to it eventually...
Regards
Henrik
mån 2008-04-07 klockan 13:58 +0800 skrev Adrian Chadd:
The first snippet of refactoring, which has been committed:
http://www.squid-cache.org/Versions/v2/HEAD/changesets/12032.patch
Yes...
Shuffling the storeClientReadHeaders() stuff out of client_side.c and into
store_client.c in
sön 2008-04-06 klockan 18:39 -0600 skrev Alex Rousskov:
That's comforting. Do any bigger-than-usual headers end up being sent to
ICAP after your patch is applied?
Yes, that may happen in trunk already, but it later gets confused in
client_side_reply.cc without the last submission.
In 3.0 we
mån 2008-04-07 klockan 16:01 +0800 skrev Adrian Chadd:
I'm trying to avoid doing the much simpler solution and ignoring the rest -
ie, just having a local buffer, reading the reply from the server-side into it
until we have all the headers, and then just copying the first 4k out.
Thats a
mån 2008-04-07 klockan 07:17 +0200 skrev Kinkie:
rsync://squid-cache.org/source/squid-version
where version is one of 3, 3.0, 2, 2.7, 2.6
I've added documentation for this on the wiki in DeveloperResources.
Thanks
Regards
Henrik
mån 2008-04-07 klockan 16:25 +0800 skrev Adrian Chadd:
You've tested 2.6 too? I read the code as handling a HIT but just not parsing
the reply and sending it back verbatim.
I didn't test 2.6 today, but I know from earlier that it falls back on
miss on large response headers.
Testing again to
mån 2008-04-07 klockan 05:02 -0400 skrev Tsantilas Christos:
This test sometimes gives me an assertion in pconn.cc (assert(index = 0)
, pconn.cc:89 line), possibly when a connection to a web server expired.
Maybe it is not related with your patch, but I am not getting this error
using normal
mån 2008-04-07 klockan 06:06 -0400 skrev Tsantilas Christos:
It is easy. I just opened 3-4 pages to wikipedia.org, some clicks on
links on those pages and then waiting for some minutes. If I try to go to
a new wikipedia page I am getting the assertion..
Seems to work fine here...
mån 2008-04-07 klockan 06:58 -0400 skrev Tsantilas Christos:
Maybe the the assertion happens to a ICAP related pconn. Give me some
time, I will repeat the tests today at the evening.
Ok.
Regards
Henrik
501 - 600 of 2137 matches
Mail list logo