You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@httpd.apache.org by Ruediger Pluem <rp...@apache.org> on 2015/08/21 16:15:22 UTC

Re: svn commit: r1696960 - in /httpd/httpd/trunk: CHANGES docs/log-message-tags/next-number modules/proxy/mod_proxy_balancer.c


On 08/21/2015 02:34 PM, ylavic@apache.org wrote:
> Author: ylavic
> Date: Fri Aug 21 12:34:02 2015
> New Revision: 1696960
> 
> URL: http://svn.apache.org/r1696960
> Log:
> mod_proxy_balancer: Fix balancers and balancer members reuse on
> restart when new ones are added.  PR 58024.
> 
> Since slotmem_create() issues a strict check on the size of an existing
> slot before reusing it, it won't reuse existing balancers and members when
> new ones are added during restart (whereas growth margins would allow it).
> Fix this by using slotmem_attach() first and if it succeeds do the checks
> based on the returned previous number of existing entries.
> 
> Modified:
>     httpd/httpd/trunk/CHANGES
>     httpd/httpd/trunk/docs/log-message-tags/next-number
>     httpd/httpd/trunk/modules/proxy/mod_proxy_balancer.c
> 
> Modified: httpd/httpd/trunk/CHANGES
> URL: http://svn.apache.org/viewvc/httpd/httpd/trunk/CHANGES?rev=1696960&r1=1696959&r2=1696960&view=diff
> ==============================================================================
> --- httpd/httpd/trunk/CHANGES [utf-8] (original)
> +++ httpd/httpd/trunk/CHANGES [utf-8] Fri Aug 21 12:34:02 2015
> @@ -1,6 +1,9 @@
>                                                           -*- coding: utf-8 -*-
>  Changes with Apache 2.5.0
>  
> +  *) mod_proxy_balancer: Fix balancers and balancer members reuse on
> +     restart when new ones are added.  PR 58024.  [Yann Ylavic]
> +
>    *) mod_socache_memcache: Add the 'MemcacheConnTTL' directive to control how 
>       long to keep idle connections with the memcache server(s).
>       Change default value from 600 usec (!) to 15 sec. PR 58091
> 
> Modified: httpd/httpd/trunk/docs/log-message-tags/next-number
> URL: http://svn.apache.org/viewvc/httpd/httpd/trunk/docs/log-message-tags/next-number?rev=1696960&r1=1696959&r2=1696960&view=diff
> ==============================================================================
> --- httpd/httpd/trunk/docs/log-message-tags/next-number (original)
> +++ httpd/httpd/trunk/docs/log-message-tags/next-number Fri Aug 21 12:34:02 2015
> @@ -1 +1 @@
> -2964
> +2966
> 
> Modified: httpd/httpd/trunk/modules/proxy/mod_proxy_balancer.c
> URL: http://svn.apache.org/viewvc/httpd/httpd/trunk/modules/proxy/mod_proxy_balancer.c?rev=1696960&r1=1696959&r2=1696960&view=diff
> ==============================================================================
> --- httpd/httpd/trunk/modules/proxy/mod_proxy_balancer.c (original)
> +++ httpd/httpd/trunk/modules/proxy/mod_proxy_balancer.c Fri Aug 21 12:34:02 2015
> @@ -761,8 +761,11 @@ static int balancer_post_config(apr_pool
>          char *id;
>          proxy_balancer *balancer;
>          ap_slotmem_type_t type;
> +        apr_size_t attached_size;
> +        unsigned int attached_num;
>          void *sconf = s->module_config;
>          conf = (proxy_server_conf *)ap_get_module_config(sconf, &proxy_module);
> +
>          /*
>           * During create_proxy_config() we created a dummy id. Now that
>           * we have identifying info, we can create the real id
> @@ -794,11 +797,39 @@ static int balancer_post_config(apr_pool
>                           (int)ALIGNED_PROXY_BALANCER_SHARED_SIZE,
>                           (int)conf->balancers->nelts, conf->max_balancers);
>  
> -            rv = storage->create(&new, conf->id,
> -                                 ALIGNED_PROXY_BALANCER_SHARED_SIZE,
> -                                 conf->max_balancers, type, pconf);
> +            /* First try to attach() since the number of configured balancers
> +             * may have changed during restart, and we don't want create() to
> +             * fail because the overall size * number of entries is not stricly
> +             * identical to the previous run.  There may still be enough room
> +             * for this new run thanks to bgrowth margin, so if attach()
> +             * succeeds we can only check for the number of available entries
> +             * to be *greater or* equal to what we need now.  If attach() fails
> +             * we simply fall back to create().
> +             */
> +            rv = storage->attach(&new, conf->id,
> +                                 &attached_size, &attached_num,
> +                                 pconf);
> +            if (rv != APR_SUCCESS) {
> +                rv = storage->create(&new, conf->id,
> +                                     ALIGNED_PROXY_BALANCER_SHARED_SIZE,
> +                                     conf->max_balancers, type, pconf);
> +            }
> +            else {
> +                ap_log_error(APLOG_MARK, APLOG_DEBUG, 0, s, APLOGNO(02964)
> +                             "Balancers attached: %d, %d (%d)",
> +                             (int)ALIGNED_PROXY_BALANCER_SHARED_SIZE,
> +                             (int)attached_num, conf->max_balancers);
> +                if (attached_size == ALIGNED_PROXY_BALANCER_SHARED_SIZE
> +                        && attached_num >= conf->balancers->nelts) {
> +                    conf->max_balancers = attached_num;
> +                }
> +                else {
> +                    rv = APR_ENOSPC;

Hm. Doesn't that mean that if I keep on adding stuff and do graceful restarts that at one point in time I will have used
up all my growth and need to do a hard restart?
If yes, it should be probably documented to avoid some unexpected surprises.


Regards

RĂ¼diger

Re: svn commit: r1696960 - in /httpd/httpd/trunk: CHANGES docs/log-message-tags/next-number modules/proxy/mod_proxy_balancer.c

Posted by Yann Ylavic <yl...@gmail.com>.
On Fri, Aug 21, 2015 at 8:07 PM, Ruediger Pluem <rp...@apache.org> wrote:
>
>
> On 08/21/2015 07:02 PM, Yann Ylavic wrote:
>> On Fri, Aug 21, 2015 at 4:15 PM, Ruediger Pluem <rp...@apache.org> wrote:
>>>
>>> Hm. Doesn't that mean that if I keep on adding stuff and do graceful restarts that at one point in time I will have used
>>> up all my growth and need to do a hard restart?
>>
>> Yes, the last restart would fail.
>>
>>> If yes, it should be probably documented to avoid some unexpected surprises.
>>
>> I agree it should be documented (will do), currently we already have
>> this issue from the first restart though.
>
> So it is an improvement over what we have :-)

Yes, including the doc :)
Added in r1697352, and to the backport proposal.

Thanks,
Yann.

Re: svn commit: r1696960 - in /httpd/httpd/trunk: CHANGES docs/log-message-tags/next-number modules/proxy/mod_proxy_balancer.c

Posted by Ruediger Pluem <rp...@apache.org>.

On 08/21/2015 07:02 PM, Yann Ylavic wrote:
> On Fri, Aug 21, 2015 at 4:15 PM, Ruediger Pluem <rp...@apache.org> wrote:
>>
>> Hm. Doesn't that mean that if I keep on adding stuff and do graceful restarts that at one point in time I will have used
>> up all my growth and need to do a hard restart?
> 
> Yes, the last restart would fail.
> 
>> If yes, it should be probably documented to avoid some unexpected surprises.
> 
> I agree it should be documented (will do), currently we already have
> this issue from the first restart though.

So it is an improvement over what we have :-)

Regards

RĂ¼diger


Re: svn commit: r1696960 - in /httpd/httpd/trunk: CHANGES docs/log-message-tags/next-number modules/proxy/mod_proxy_balancer.c

Posted by Yann Ylavic <yl...@gmail.com>.
On Fri, Aug 21, 2015 at 4:15 PM, Ruediger Pluem <rp...@apache.org> wrote:
>
> Hm. Doesn't that mean that if I keep on adding stuff and do graceful restarts that at one point in time I will have used
> up all my growth and need to do a hard restart?

Yes, the last restart would fail.

> If yes, it should be probably documented to avoid some unexpected surprises.

I agree it should be documented (will do), currently we already have
this issue from the first restart though.

Regards,
Yann.