On Tue, Apr 13, 2021 at 10:52:15PM -0600, Thayne McCombs wrote:
>
> On 4/13/21 3:44 AM, Willy Tarreau wrote:
> > I'm failing to see how it differs from regsub() which already does the
> > same with the reference (you'd write \1 instead of 1) and also allows to
> > compose something out of multiple
On 4/13/21 3:44 AM, Willy Tarreau wrote:
I'm failing to see how it differs from regsub() which already does the
same with the reference (you'd write \1 instead of 1) and also allows to
compose something out of multiple matches. Am I missing something, or a
case where regsub() is really not conv
On Wed, Apr 14, 2021 at 03:02:20AM +0200, Aleksandar Lazic wrote:
> > But then, could it make sense to also support "strict integers": values
> > that can accurately be represented as integers and which are within the
> > JSON valid range for integers (-2^52 to 2^52 with no decimal part) ?
> > This
On Wed, Apr 14, 2021 at 02:03:52AM +, Rmrf99 wrote:
> Thanks Chris, Willy! this make me clear now.
> Glad to see dynamically add/remove servers feature under development.
If you're interested by with feature, please try the latest 2.4-dev,
you already have the "add server" feature on the comma
Thanks Chris, Willy! this make me clear now.
Glad to see dynamically add/remove servers feature under development.
‐‐‐ Original Message ‐‐‐
On Wednesday, April 14, 2021 1:19 AM, Christopher Faulet
wrote:
> Le 13/04/2021 à 18:15, John Lauro a écrit :
>
> > Sounds like the biggest part o
On 13.04.21 11:26, Willy Tarreau wrote:
Hi Aleks,
On Mon, Apr 12, 2021 at 10:09:08PM +0200, Aleksandar Lazic wrote:
Hi.
another patch which honer the feedback.
Thank you. FWIW I agree with all the points reported by Tim. I'll add
a few comments and/or suggestions below. On a general note, pl
Hi,
This is a friendly bot that watches fixes pending for the next haproxy-stable
release! One such e-mail is sent periodically once patches are waiting in the
last maintenance branch, and an ideal release date is computed based on the
severity of these fixes and their merge date. Responses t
Le 13/04/2021 à 18:15, John Lauro a écrit :
Sounds like the biggest part of hot restarts is the cost of leaving the old
process running as they have a lot of long running TCP connections, and if you
do a lot of restarts the memory requirements build up. Not much of an issue for
short lived htt
On Tue, Apr 13, 2021 at 12:15:33PM -0400, John Lauro wrote:
> Sounds like the biggest part of hot restarts is the cost of leaving the old
> process running as they have a lot of long running TCP connections, and if
> you do a lot of restarts the memory requirements build up. Not much of an
> issue
Le 13/04/2021 à 18:05, Tim Düsterhus a écrit :
Christopher,
On 4/13/21 4:59 PM, Christopher Faulet wrote:
+/* Sorts the parameters within the given query string. Returns an ist
containing
+ * the new path and backed by `trash` or IST_NULL if the `len` not
sufficiently
+ * large to store the res
Le 13/04/2021 à 18:03, Tim Düsterhus a écrit :
Christopher,
On 4/13/21 4:38 PM, Christopher Faulet wrote:
At the end it remains your choice. The function is quite good. I just
wonder if it could be valuable to also handle single dot-segment here in
addition to double dot-segment. Thus, the norm
Le 13/04/2021 à 17:45, Tim Düsterhus a écrit :
Christopher,
On 4/13/21 2:41 PM, Christopher Faulet wrote:
Sorry for the delay. I'll comment your patches by replying inline when
No delay experienced. You said that you'd try this week and it's still
this week. So this is fine :-)
appropriate.
Sounds like the biggest part of hot restarts is the cost of leaving the old
process running as they have a lot of long running TCP connections, and if
you do a lot of restarts the memory requirements build up. Not much of an
issue for short lived http requests (although it would be nice if keep
al
Le 08/04/2021 à 20:59, Tim Duesterhus a écrit :
Willy,
Christopher,
and this final one adds a normalizer to turn the hex digits of percent
encoding into uppercase. Uppercase is the variant preferred by the URI RFC, so
this is what we do.
This one looks good.
--
Christopher Faulet
Christopher,
On 4/13/21 4:59 PM, Christopher Faulet wrote:
+/* Sorts the parameters within the given query string. Returns an ist
containing
+ * the new path and backed by `trash` or IST_NULL if the `len` not
sufficiently
+ * large to store the resulting path.
+ */
+struct ist uri_normalizer_q
Christopher,
On 4/13/21 4:38 PM, Christopher Faulet wrote:
At the end it remains your choice. The function is quite good. I just
wonder if it could be valuable to also handle single dot-segment here in
addition to double dot-segment. Thus, the normalizer should be renamed
"dot-segments" or som
Le 08/04/2021 à 20:59, Tim Duesterhus a écrit :
Willy,
Christopher,
this is in prepatation for the next normalizer which normalizes character case
of the percent encoding.
The resources I found are not clear on whether a percent that is not followed
by two hex digits is valid or not. Most brows
Christopher,
On 4/13/21 2:41 PM, Christopher Faulet wrote:
Sorry for the delay. I'll comment your patches by replying inline when
No delay experienced. You said that you'd try this week and it's still
this week. So this is fine :-)
appropriate. The quality of the whole series is pretty good
On Tue, Apr 13, 2021 at 04:44:38PM +0200, Moemen MHEDHBI wrote:
> > But then how about having just *your* functions
> > without relying on the other ones ? Now that you've extended the existing
> > function, you can declare it yours, remove all the configurable stuff and
> > keep the simplified ver
Le 08/04/2021 à 20:59, Tim Duesterhus a écrit :
Willy,
Christopher,
most of the patch is moving around the config parser to support ingesting the
new argument.
This one looks good.
--
Christopher Faulet
Le 08/04/2021 à 20:59, Tim Duesterhus a écrit :
Willy,
Christopher,
I did not perform any measurements at all. But not reallocating for every
parameter should be better :-)
This one may be useless if you use the trash buffer to store the query
parameters.
--
Christopher Faulet
Le 08/04/2021 à 20:59, Tim Duesterhus a écrit :
Willy,
Christopher,
This one comes with dynamic allocation. The next patch will add an optimization
for a small number of arguments. However dynamic allocation within the main
processing logic is pretty ugly, so this should be looked at further.
On 13/04/2021 11:39, Willy Tarreau wrote:
>> You can find attached the patches 0001-bis and 0002-bis modifying the
>> existing functions (introducing an url flag) to see how it looks like.
>> This solution may be cleaner (no chunk allocation and we don't loop
>> twice over input string) but has t
Le 08/04/2021 à 20:59, Tim Duesterhus a écrit :
Willy,
Christopher,
I'm not very happy with the normalization logic, because am processing the URI
in reverse. This requires me to directly access offsets instead of using the
`ist` API. However this way I don't need to backtrack once I encounter a
Le 08/04/2021 à 20:59, Tim Duesterhus a écrit :
Willy,
Christopher,
something simple for a start. This one adds the http-request action and a
very simple normalizer to test whether it works. Turns out it does :-)
You can see the new `ist` helpers in action already. I'm pretty happy that
I was a
Le 08/04/2021 à 20:59, Tim Duesterhus a écrit :
Willy,
Christopher,
I used uri_auth.[ch] as the basis for the source file structure (comments and
stuff).
Thanks, nothing to say about this one :)
--
Christopher Faulet
Le 08/04/2021 à 20:59, Tim Duesterhus a écrit :
Willy,
Christopher,
Not sure who of you is better suited to review this series, so I'm adding both
of you :-)
I'm tagging this as RFC, because it's large and quite a bit outside of my
comfort zone. However the patches are as clean as possible. The
Hi list
Moving from httpd (reverse proxy) to haproxy 2.3.9
In httpd i have following:
ServerName test.server.dk
ProxyPreserveHost On
ProxyPass / http://localhost:8080/aaa/bbb/
ProxyPassReverse / http://localhost:8080/aaa/bbb/
What is the correct way to do this in haproxy 2 ?
Something with http
On Tue, Apr 13, 2021 at 01:31:12AM +, Rmrf99 wrote:
> In this Slack engineering blog post:
> https://slack.engineering/migrating-millions-of-concurrent-websockets-to-envoy/
>
> they replace HAProxy with Envoy for **Hot Restart**, just curious does
> HAProxy new version will have similar appro
On Sat, Apr 10, 2021 at 01:34:16PM +0200, Ionel GARDAIS wrote:
> Hi team, list,
>
> If you haven't already read it, go read this blog article :
> [
> https://www.haproxy.com/blog/haproxy-forwards-over-2-million-http-requests-per-second-on-a-single-aws-arm-instance/
> |
> https://www.haproxy.c
Hi Thayne,
On Sun, Apr 11, 2021 at 11:26:59PM -0600, Thayne McCombs wrote:
> Adjust the size of the sample buffer before we change the "area"
> pointer. The change in size is calculated as the difference between the
> original pointer and the new start pointer. But since the
> `smp->data.u.str.are
Hi Thayne,
On Tue, Apr 13, 2021 at 02:11:25AM -0600, Thayne McCombs wrote:
> Add a new sample converter that finds the first regex match and returns
> the substring for that match, or a capture group, if an index is
> provided.
> ---
> doc/configuration.txt| 22 +++
> reg-test
Hi Jarno,
On Tue, Apr 13, 2021 at 06:19:47AM +, Jarno Huuskonen wrote:
> Hello,
>
> On Tue, 2021-04-06 at 01:58 +0200, Moemen MHEDHBI wrote:
> > Thanks Willy and Tim for your feedback.
> >
> > You can find attached the updated patches with fixed coding style (now
> > set correctly in my edit
Hi Moemen,
On Tue, Apr 13, 2021 at 12:41:39AM +0200, Moemen MHEDHBI wrote:
> >> in such case should we rather use dynamic allocation ?
> >
> > No, there are two possible approaches. One of them is to use a trash
> > buffer using get_trash_chunk(). The trash buffers are "large enough"
> > for anyth
Hi Aleks,
On Mon, Apr 12, 2021 at 10:09:08PM +0200, Aleksandar Lazic wrote:
> Hi.
>
> another patch which honer the feedback.
Thank you. FWIW I agree with all the points reported by Tim. I'll add
a few comments and/or suggestions below. On a general note, please be
careful about your indenting,
NFP WORKSHOPS
Affordable Training Courses
Bid Writing: The Basics
Do you know the most common reasons for rejection? Are you gathering the right
evidence? Are you making the right arguments? Are you using the right
terminology? Are your numbers right? Are you learning from rejections?
Add a new sample converter that finds the first regex match and returns
the substring for that match, or a capture group, if an index is
provided.
---
doc/configuration.txt| 22 +++
reg-tests/converter/regmatch.vtc | 39 +++
src/sample.c | 66
37 matches
Mail list logo
