Re: Set State to DRAIN vs set weight 0

On Mon, Jan 25, 2016 at 12:50:48AM +0100, Baptiste wrote: > That said, I may have found something in accurate. We'll discuss about > it tomorrow in the office. OK. > > BTW, maybe we should provide this cmd line option, then it will be possible > > to state that a service reload maintain the

Re: ssl performance regression in version 1.6

Hi again, On Sun, Jan 24, 2016 at 09:00:31PM +0100, Willy Tarreau wrote: > OK so I'm having a preliminary patch for the performance regression > causing incomplete buffers to be processed. It's not optimal yet, > but it's safe and works for me. Now I can reproduce the tests and > always observe

Re: ssl performance regression in version 1.6

Hi Cyril, On Sat, Jan 23, 2016 at 11:55:27PM +0100, Cyril Bonté wrote: > From my first tests, I can see a performance drop between commit > 5506e3f8 and e583ea58 (before 1.6-dev1). > > Here is the sample configuration I use : > global > tune.ssl.default-dh-param 2048 > > listen test >

Re: Temporary Maintenance frontend for all port 80/443 ?

The way we've done it has been to add a maintenance server to all of our backends, then use the admin socket to iteratively disable all "real" servers for our backends. The maintenance server points all requests to your temporary page. Something like: server maintenance 10.0.5.10:80 backup

Re: Temporary Maintenance frontend for all port 80/443 ?

We've tended to disable backend servers for specific backends as part of our deployer integration, so it makes a certain amount of sense for us to use the same basic methodology to push something into a particular backend. (We also don't have to restart haproxy to adjust the configuration,

Temporary Maintenance frontend for all port 80/443 ?

Hi, I wondered if it's possible, I seen some examples but was not sure yet, to let HA proxy running as you do but place a frontend that listens on 0.0.0.0 and port 80 and 443 and redirect all requests to a temporary page ? Maybe even a local page on HAproxy ? As I'm on pfsense I need to disable

Re: Temporary Maintenance frontend for all port 80/443 ?

Am 24.01.2016 17:58 schrieb Jeff Buchbinder: The way we've done it has been to add a maintenance server to all of our backends, then use the admin socket to iteratively disable all "real" servers for our backends. The maintenance server points all requests to your temporary page. Something

Re: ssl performance regression in version 1.6

On Sun, Jan 24, 2016 at 07:51:16PM +0100, Cyril Bonté wrote: > Le 24/01/2016 19:42, Willy Tarreau a écrit : > >- when I tried to lower the buffer size, I entered exactly 16000. For a > > reason I ignore, the transfer hangs after the end of the first > > transfer. > > I have not investigated

Re: ssl performance regression in version 1.6

Hi Willy, Le 24/01/2016 18:03, Willy Tarreau a écrit : Hi Cyril, On Sat, Jan 23, 2016 at 11:55:27PM +0100, Cyril Bonté wrote: From my first tests, I can see a performance drop between commit 5506e3f8 and e583ea58 (before 1.6-dev1). Here is the sample configuration I use : global

Re: ssl performance regression in version 1.6

Le 24/01/2016 19:42, Willy Tarreau a écrit : - when I tried to lower the buffer size, I entered exactly 16000. For a reason I ignore, the transfer hangs after the end of the first transfer. I have not investigated this and it could be my server (a hacked haproxy 1.6 modified to

Re: ssl performance regression in version 1.6

On Sun, Jan 24, 2016 at 07:29:36PM +0100, Cyril Bonté wrote: > I'm also testing on my laptop on localhost to avoid some network side > effects. Here, the tests are consistent, I always have the same > performance for a given version. OK fine. > >Under stace, it's even worse. The number of

Re: Set State to DRAIN vs set weight 0

On Fri, Jan 22, 2016 at 10:46 AM, Willy Tarreau wrote: > Hi Alex, > > On Fri, Jan 22, 2016 at 11:32:14AM +0200, Alex wrote: >> Hi, >> >> Thank you for the answer, this is very helpful. >> So to sum up my understanding, usually drain is used in operations by >> setting the server in

Re: Set State to DRAIN vs set weight 0

Hi Baptiste, On Sun, Jan 24, 2016 at 09:01:27PM +0100, Baptiste wrote: > I commented this case in the code: > > /* apply drain mode if server is currently enabled */ > if (!(srv->admin & SRV_ADMF_FMAINT) && (srv_admin_state & SRV_ADMF_FDRAIN)) { > /* The SRV_ADMF_FDRAIN flag is

Re: Set State to DRAIN vs set weight 0

On Sun, Jan 24, 2016 at 9:01 PM, Baptiste wrote: > On Fri, Jan 22, 2016 at 10:46 AM, Willy Tarreau wrote: >> Hi Alex, >> >> On Fri, Jan 22, 2016 at 11:32:14AM +0200, Alex wrote: >>> Hi, >>> >>> Thank you for the answer, this is very helpful. >>> So to sum up my

Re: Set State to DRAIN vs set weight 0

On Mon, Jan 25, 2016 at 12:35 AM, Willy Tarreau wrote: > Hi Baptiste, > > On Sun, Jan 24, 2016 at 09:01:27PM +0100, Baptiste wrote: >> I commented this case in the code: >> >> /* apply drain mode if server is currently enabled */ >> if (!(srv->admin & SRV_ADMF_FMAINT) &&