On Wed, Apr 14, 2021 at 07:16:25PM +0200, Tim Düsterhus wrote:
> Willy,
>
> On 4/14/21 7:00 PM, ??? wrote:
> > something changed in freebsd packaging. we now need to install pcre
> > directly.
>
> This one looks good to me.
Applied, thanks!
Willy
On Wed, Apr 14, 2021 at 09:52:31PM +0200, Aleksandar Lazic wrote:
> > > + - string : This is the default search type and returns a String;
> > > + - boolean : If the JSON value is not a String or a Number
> > > + - number : When the JSON value is a Number then will the value be
> > > +
On Thu, Apr 15, 2021 at 09:02:18AM +0200, Ferry Bodijn wrote:
> L.S.
>
> Is it possible with HAProxy, to have a frontend which binds on HTTP that
> refers to a backend that connects to an HTTPS endpoint using Client
> Certificate Authentication while reaching it via a Squid forwarding proxy?
If
On Thu, Apr 15, 2021 at 08:59:35AM +0200, Willy Tarreau wrote:
> On Wed, Apr 14, 2021 at 01:53:06PM +0200, Christopher Faulet wrote:
> > > nbthread=64, nbproc=1 on both 1.8/2.x
> >
> > It is thus surprising, if it is really a contention issue, that you never
> > observed slow down on the 1.8.
On Wed, Apr 14, 2021 at 01:53:06PM +0200, Christopher Faulet wrote:
> > nbthread=64, nbproc=1 on both 1.8/2.x
>
> It is thus surprising, if it is really a contention issue, that you never
> observed slow down on the 1.8. There is no watchdog, but the thread
> implementation is a bit awkward on
L.S.
Is it possible with HAProxy, to have a frontend which binds on HTTP that
refers to a backend that connects to an HTTPS endpoint using Client
Certificate Authentication while reaching it via a Squid forwarding proxy?
I have a working Apache HTTP server config which I want to replace because
Migrating from stock stretch-backports+1.8 to Debian_10/Buster+2.0 (to
purge 'reqrep' en route to 2.2), /var/log/ is suddenly filling disk
with messages about changed IPs :
===
2021-04-14T01:08:40.123303+00:00 ip-10-36-217-169 haproxy[569]:
my_web_service changed its IP from 52.208.198.117 to
Hello,
On Thu, 2021-04-15 at 01:43 -0600, Jim Freeman wrote:
> This is puzzling, since haproxy.cfg directs all logs to local*
> After some investigation, it turns out that the daemon.log and syslog
> entries arrive via facility.level=daemon.info. I've made rsyslog cfg
> changes that now stop the
On Thu, Apr 15, 2021 at 07:13:53AM +, Robin H. Johnson wrote:
> Thanks; I will need to catch it faster or automate this, because the
> watchdog does a MUCH better job restarting it than before, less than 30
> seconds of 100% CPU before the watchdog reliably kills it.
I see. Then collecting
On Wed, Apr 14, 2021 at 12:56:39PM +0200, Miroslav Zagorac wrote:
> Hello all,
>
> the next two patches allow you to transfer the OpenTracing context via an
> HTTP headers without using a name prefix.
>
> This is very useful when transferring context from or to some other process
> that cannot
Hi Robert,
On Wed, Apr 14, 2021 at 12:18:42PM +0200, Robert Ionescu wrote:
> Hi all,
>
> I am working on a new patch for haproxy and would need the malloc method
> used to allocate memory for structs like the connection struct in
> connection-t.c
> In case I want to expand the struct with a new
Great work,
i testet it right now and work like a charm!
Thank u Zaga, good work!
Il giorno gio 15 apr 2021 alle ore 08:40 Willy Tarreau ha
scritto:
> On Wed, Apr 14, 2021 at 12:56:39PM +0200, Miroslav Zagorac wrote:
> > Hello all,
> >
> > the next two patches allow you to transfer the
On 15.04.21 09:08, Willy Tarreau wrote:
On Wed, Apr 14, 2021 at 09:52:31PM +0200, Aleksandar Lazic wrote:
+ - string : This is the default search type and returns a String;
+ - boolean : If the JSON value is not a String or a Number
+ - number : When the JSON value is a Number then will
On 15.04.21 17:09, Willy Tarreau wrote:
On Thu, Apr 15, 2021 at 04:49:00PM +0200, Aleksandar Lazic wrote:
#define JSON_INT_MAX ((1ULL << 53) - 1)
^
Sorry I was not clear, please drop that 'U' here.
I'm also sorry, I was in a tunnel :-/
Attached now the next
On Thu, Apr 15, 2021 at 04:49:00PM +0200, Aleksandar Lazic wrote:
> > > #define JSON_INT_MAX ((1ULL << 53) - 1)
> >^
> > Sorry I was not clear, please drop that 'U' here.
>
> I'm also sorry, I was in a tunnel :-/
>
> Attached now the next patches.
Thank you! Now
On Thu, Apr 15, 2021 at 07:19:25PM +0500, ??? wrote:
> Hello,
>
> as described in travis-ci docs:
> https://blog.travis-ci.com/2020-09-11-arm-on-aws
>
> it is next generation armv8 available on AWS.
Applied, thanks Ilya!
Willy
Willy,
On 4/15/21 5:09 PM, Willy Tarreau wrote:
On Thu, Apr 15, 2021 at 04:49:00PM +0200, Aleksandar Lazic wrote:
#define JSON_INT_MAX ((1ULL << 53) - 1)
^
Sorry I was not clear, please drop that 'U' here.
I'm also sorry, I was in a tunnel :-/
Attached now the
On Thu, Apr 15, 2021 at 02:17:45PM +0200, Aleksandar Lazic wrote:
> > I, by far, prefer Tim's proposal here, as I do not even understand the
> > first one, sorry Aleks, please don't feel offended :-)
>
> Well you know my focus is to support HAProxy and therefore it's okay.
> The contribution was
On 15.04.21 14:48, Willy Tarreau wrote:
On Thu, Apr 15, 2021 at 02:17:45PM +0200, Aleksandar Lazic wrote:
I, by far, prefer Tim's proposal here, as I do not even understand the
first one, sorry Aleks, please don't feel offended :-)
Well you know my focus is to support HAProxy and therefore
On Thu, Apr 15, 2021 at 03:41:18PM +0200, Aleksandar Lazic wrote:
> > > Now when I remove the check "smp->data.u.sint < 0" every positive value is
> > > bigger then JSON INT_MIN and returns 0.
> >
> > But don't you agree that this test DOES nothing ? If it changes anything
> > it means the issue
Hi all,
is it possible to use a variable as span name?
for example, can i use uri as span name so in zipkin it's visible as
service:
Thanks
--
Bonini Andrea
On 04/15/2021 04:24 PM, Andrea Bonini wrote:
Hi all,
is it possible to use a variable as span name?
for example, can i use uri as span name so in zipkin it's visible as
service:
Thanks
Hello Andrea,
the name of the span is defined statically and cannot be the content of
a variable.
On 15.04.21 15:55, Willy Tarreau wrote:
On Thu, Apr 15, 2021 at 03:41:18PM +0200, Aleksandar Lazic wrote:
Now when I remove the check "smp->data.u.sint < 0" every positive value is
bigger then JSON INT_MIN and returns 0.
But don't you agree that this test DOES nothing ? If it changes anything
Hello,
as described in travis-ci docs:
https://blog.travis-ci.com/2020-09-11-arm-on-aws
it is next generation armv8 available on AWS.
cheers,
Ilya
From 4365c936396a5c7c4f710ed8267b06a73fd5bbcc Mon Sep 17 00:00:00 2001
From: Ilya Shipitsin
Date: Thu, 15 Apr 2021 19:16:09 +0500
Subject: [PATCH]
On 15.04.21 16:09, Willy Tarreau wrote:
On Thu, Apr 15, 2021 at 04:05:27PM +0200, Aleksandar Lazic wrote:
Well I don't think so because 4 is still bigger then -9007199254740991 ;-)
This is because *you* think it is -9007199254740991 but the reality
is that it's not this.due to ULL:
On Thu, Apr 15, 2021 at 04:05:27PM +0200, Aleksandar Lazic wrote:
> Well I don't think so because 4 is still bigger then -9007199254740991 ;-)
This is because *you* think it is -9007199254740991 but the reality
is that it's not this.due to ULL:
#define JSON_INT_MAX ((1ULL << 53) - 1)
#define
On Thu, Apr 15, 2021 at 09:23:07AM +0200, Willy Tarreau wrote:
> On Thu, Apr 15, 2021 at 07:13:53AM +, Robin H. Johnson wrote:
> > Thanks; I will need to catch it faster or automate this, because the
> > watchdog does a MUCH better job restarting it than before, less than 30
> > seconds of
Christopher,
this one should be unchanged. I just fixed the conflict with Aleks' JSON patch.
Best regards
Tim Düsterhus
Apply with `git am --scissors` to automatically cut the commit message.
-- >8 --
This is in preparation for future patches.
---
Makefile | 2 +-
Christopher,
this one is completely new. It cleanly separates the patches for adding the
action from the patches adding the normalizers.
Best regards
Tim Düsterhus
Apply with `git am --scissors` to automatically cut the commit message.
-- >8 --
This patch adds the `http-request normalize-uri`
Christopher,
the general logic of the normalizer is unchanged, but the whole framework was
refactored.
Best regards
Tim Düsterhus
Apply with `git am --scissors` to automatically cut the commit message.
-- >8 --
This normalizer merges adjacent slashes into a single slash, thus removing
empty
Christopher,
in this one I cleaned up the values of the enum.
Best regards
Tim Düsterhus
Apply with `git am --scissors` to automatically cut the commit message.
-- >8 --
This enum will serve as the return type for each normalizer.
---
include/haproxy/uri_normalizer-t.h | 31
Christopher,
the general logic of the normalizer is unchanged, but the whole framework was
refactored.
Best regards
Tim Düsterhus
Apply with `git am --scissors` to automatically cut the commit message.
-- >8 --
This normalizer merges `../` path segments with the predecing segment, removing
Christopher,
the general logic of the normalizer is unchanged, but the whole framework was
refactored.
Best regards
Tim Düsterhus
Apply with `git am --scissors` to automatically cut the commit message.
-- >8 --
This normalizer uppercases the hexadecimal characters used in percent-encoding.
Christopher,
here I am using `istdiff` and a trash buffer instead of dynamic allocation.
Best regards
Tim Düsterhus
Apply with `git am --scissors` to automatically cut the commit message.
-- >8 --
This normalizer sorts the `&` delimited query parameters by parameter name.
See GitHub Issue
Christopher,
again: Thank you for the very helpful review. In this series you can find v2 of
my URI normalization patches. I hope I did not forget to incorporate any of
your feedback.
Some general notes:
- I completely cleaned up the commit history to group similar patches (e.g. the
two
Christopher,
the general logic of the normalizer is unchanged, but the whole framework was
refactored.
Best regards
Tim Düsterhus
Apply with `git am --scissors` to automatically cut the commit message.
-- >8 --
This adds an option to supress `../` at the start of the resulting path.
---
Yes - as a systemd service. But the puzzlement remains that the same complaints
get logged via *both* daemon.info and local*.info, when local* is all
we configure
in haproxy.cfg. /var/log gets doubly over-taxed because the daemon.info entries
wind up in both syslog and daemon.log (per stock
On Thu, Apr 15, 2021 at 07:53:15PM +, Robin H. Johnson wrote:
> But your thought of CPU pinning was good.
> I went to confirm it in the host, and I'm not certain if the cpu-map is
> working
> right.
Ignore me, long day and I didn't think to check each thread PID:
# ps -e -T | grep haproxy -w
More info on the over-quick+newly-noisy resolves that were triggering this ...
We've been running 1.8.19
(https://packages.debian.org/stretch-backports/haproxy)
with 'hold valid 60s' configured, which was acting-ish like 'timeout
resolve 60s'
(which was *not* configured).
So when we moved to
39 matches
Mail list logo
