Re: Block clients based on header in real time?

Hello, Pass these time, i return to this situation. I try to implement in this stick table a white and black list, one solution is based on storing the ips  and play with setting data.gpc0 1 or 0, ok, it work, but the problem is now with networks. The first isue is with the stick-table, this

init variable from ACL condition

hello, i would like to change the backend's port dynamically, is it possible to do something like: frontend main acl myport=3000 path_beg -i /uri_one acl myport=3001 path_beg -i /uri_two backend back server web1 localhost:${myport} today, for this purpose i have to create

stick-table filled with wrong server_id?

Hi, I observed that stick store-response set-cookie is not always using the server_id of the backend-server actually setting this cookie. The backend with server_id=2 was changed that way, that it adds gf22 to the cookie-value. Following have been found in the stick table: 0x15044b4:

Enjoy High Speed Maxis fibre Internet 10mbps only at RM148/month!

Title: MAXIS HOME AND BUSINESS FIBRE INTERNET MAXIS HOME AND BUSINESS FIBRE INTERNET Fastest Fibre Broadband in Malaysia with lowest price 10Mbps, 20Mbps, & 30Mbps from only RM148

is balance uri a form of stickiness?

I've got a backend where the balancing is done via url hash (consistent.) I set default-server maxconn to 4 and I notice that when one of the servers hit it's maxconn, it queued requests on the server and didn't get redistributed to other servers in the backend. The backend queue has remained 0.

FIXED: Re: 100% cpu load....

Hi Mark, OK I could reproduce, debug and fix. It was a tough one, really... More a problem of internal semantics than anything else, so I had to test several possibilities and study their impacts and the corner cases. In the end we get something that's fixed and better :-) The issue was mostly

Re: is balance uri a form of stickiness?

Hi David, On Thu, Jul 18, 2013 at 11:56:13AM -0700, David Birdsong wrote: I've got a backend where the balancing is done via url hash (consistent.) I set default-server maxconn to 4 and I notice that when one of the servers hit it's maxconn, it queued requests on the server and didn't get

Re: Q: Best practice for redirects

Hi guys, On Thu, Jul 18, 2013 at 09:18:48AM +0200, Jürgen Haas wrote: Hi Lukas, this is great, didn't know that there is a shortcut for ACL written in {...}. You're absolutely right that this is much more readable and also I don't need those unique variable names holding the TRUE/FALSE

Re: is balance uri a form of stickiness?

On Thu, Jul 18, 2013 at 1:49 PM, Willy Tarreau w...@1wt.eu wrote: Hi David, On Thu, Jul 18, 2013 at 11:56:13AM -0700, David Birdsong wrote: I've got a backend where the balancing is done via url hash (consistent.) I set default-server maxconn to 4 and I notice that when one of the

ECC handhake failure

Hi everyone, I built haproxy with static openssl, both from latest git versions. The resulting binary works fine, but Elliptic Curve ciphers aren't available. I'm looking for help to diagnose this issue. $ openssl

Re: is balance uri a form of stickiness?

On Thu, Jul 18, 2013 at 02:37:16PM -0700, David Birdsong wrote: A tuneable would be interesting to have, but how would you envision the implementation? In the case of consistent hashing, would the request be re-routed to another server in a similar way for the case of target server being down?

Re: is balance uri a form of stickiness?

On Thu, Jul 18, 2013 at 3:12 PM, Willy Tarreau w...@1wt.eu wrote: On Thu, Jul 18, 2013 at 02:37:16PM -0700, David Birdsong wrote: A tuneable would be interesting to have, but how would you envision the implementation? In the case of consistent hashing, would the request be re-routed to

RE: ECC handhake failure

Hi Julien, $ openssl s_client -connect target:443 -cipher ECDHE-RSA-AES256-GCM-SHA384 CONNECTED(0003) 140250231854784:error:14077410:SSL routines:SSL23_GET_SERVER_HELLO:sslv3 alert handshake failure:s23_clnt.c:755: I would suggest: - try OpenSSL stable (1.0.1e) - try different ECDHE

Re: Confused by the behaviour of calloc during init() in haproxy.c

Hi Godbach, On Fri, Jul 19, 2013 at 12:21:23PM +0800, Godbach wrote: Hi Willy, In my opinion, haproxy will full size memory after allocating that for fdtab and fdinfo since calloc is used. But after I started haproxy with maxconn 1,048,576 wiht lastest snapshot, the result from `top`

Re: Confused by the behaviour of calloc during init() in haproxy.c

Hi Willy, On 2013/7/19 13:32, Willy Tarreau wrote: Hi Godbach, On Fri, Jul 19, 2013 at 12:21:23PM +0800, Godbach wrote: You should strace it. I'm sure you'll see an mmap() call that matches the size of your allocation, meaning that the libc has decided to use this instead of sbrk() to

Re: stick-table filled with wrong server_id?

*Workaround* Since I can control the cookie-value in a certain way, I can enforce my desired behaviour by instead of using stick-table doing the following: use-server ngx-gf21 if { cook(JSESSIONID) -m end gf21 } use-server ngx-gf22 if { cook(JSESSIONID) -m end gf22 } Which worked for 15.000