Please don't hesitate to push back and tell me if I can supply the patch or update in some easier-to-digest form.  In particular, while I have rigorously stress-tested this change using mod_pagespeed's unit test, system-test, and load-test framework, I don't really understand what the testing flow is for APR.  I'd be happy to add unit-tests for that if someone points me to a change-list or patch-file that does it properly.


On Thu, Nov 1, 2012 at 8:04 AM, Joshua Marantz <> wrote:
I have completed a solution to this problem, which can be a drop-in update for the existing apr_memcache.c.  It is now checked in for my module as

It differs from the solution in in that:
  • It doesn't require an API change; it but it enforces the 50ms timeout that already exists for apr_multgetp for all operations.
  • It works under my load test (which I found is not true of the patch in 51065).
For my own purposes, I will be shipping my module with apr_memcache2 so I get the behavior I want regardless of what version of Apache is installed.  But I'd like to propose my patch for apr_memcache.c.  The patch is attached, and I've also submitted it as an alternative patch to bug 51065.

If you agree with the strategy I used to solve this problem, then please let me know if I can help with any changes required to get this into the main distribution, 

On Mon, Oct 22, 2012 at 5:21 PM, Joshua Marantz <> wrote:
I've had some preliminary success with my own variant of apr_memcache.c (creatively called apr_memcache2.c).  Rather than setting the socket timeout, I've been mimicing the timeout strategy I saw in apr_memcache_multgetp, by adding a new helper method:

static apr_status_t wait_for_server_or_timeout(apr_pool_t* temp_pool,
                                               apr_memcache2_conn_t* conn) {
    apr_pollset_t* pollset;
    apr_status_t rv = apr_pollset_create(&pollset, 1, temp_pool, 0);
    if (rv == APR_SUCCESS) {
        apr_pollfd_t pollfd;
        pollfd.desc_type = APR_POLL_SOCKET;
        pollfd.reqevents = APR_POLLIN;
        pollfd.p = temp_pool;
        pollfd.desc.s = conn->sock;
        pollfd.client_data = NULL;
        apr_pollset_add(pollset, &pollfd);
        apr_int32_t queries_recvd;
        const apr_pollfd_t* activefds;
        rv = apr_pollset_poll(pollset, MULT_GET_TIMEOUT, &queries_recvd,
        if (rv == APR_SUCCESS) {
            assert(queries_recvd == 1);
            assert(activefds->desc.s == conn->sock);
            assert(activefds->client_data == NULL);
    return rv;

And calling that before many of the existing calls to get_server_line as:

    rv = wait_for_server_or_timeout_no_pool(conn);
    if (rv != APR_SUCCESS) {
        ms_release_conn(ms, conn);
        return rv;

This is just an experiment; I think I can streamline this by pre-populating the pollfd structure as part of the apr_memcache_conn_t (actually now apr_memcache2_conn_t).

I have two questions about this:

1. I noticed the version of apr_memcache.c that ships with Apache 2.4 is somewhat different from the one that ships with Apache 2.2.  In particular the 2.4 version cannot be compiled against the headers that come with a 2.2 distribution.  Is there any downside to taking my hacked 2.2 apr_memcache.c and running it in Apache 2.4?  Or should I maintain two hacks?

2. This seems wasteful in terms of system calls.  I am making an extra call to poll, rather than relying on the socket timeout.  The socket timeout didn't work as well as this though.  Does anyone have any theories as to why, or what could be done to the patch in to work?


On Fri, Oct 19, 2012 at 9:25 AM, Joshua Marantz <> wrote:
Following up: I tried doing what I suggested above: patching that change into my own copy of apr_memcache.c  It was first of all a bad idea to pull in only part of apr_memcache.c because that file changed slightly between 2.2 and 2.4 and our module works in both.

I was successful making my own version of apr_memcache (renaming entry-points apr_memcache2*) that I could hack.  But if I changed the socket timeout from -1 to 10 seconds, then the system behaved very poorly under load test (though it worked fine in our unit-tests and system-tests).  In other words, I think the proposed patch that Jeff pointed to above is not really working (as he predicted).  This test was done without SIGSTOPing the memcached; it would timeout under our load anyway and thereafter behave poorly.

I'm going to follow up on that bugzilla entry, but for now I'm going to pursue my own complicated mechanism of timing out the calls from my side.


On Thu, Oct 18, 2012 at 10:46 AM, Joshua Marantz <> wrote:
Thanks Jeff, that is very helpful.  We are considering a course of action and before doing any work toward this, I'd like to understand the pitfalls from people that understand Apache better than us.

Here's our reality: we believe we need to incorporate memcached for mod_pagespeed to scale effectively for very large sites & hosting providers.  We are fairly close (we think) to releasing this functionality as beta.  However, in such large sites, stuff goes wrong: machines crash, power failure, fiber cut, etc.  When it does we want to fall back to serving partially unoptimized sites rather than hanging the servers.

I understand the realities of backward-compatible APIs.  My expectation is that this would take years to make it into an APR distribution we could depend on.  We want to deploy this functionality in weeks.  The workarounds we have tried backgrounding the apr_memcache calls in a thread and timing out in mainline are complex and even once they work 100% will be very unsatisfactory (resource leaks; Apache refusing to exit cleanly on 'apachectl stop') if this happens more than (say) once a month.

Our plan is to copy the patched implementation of apr_memcache_server_connect and the static methods it calls into a new .c file we will link into our module, naming the new entry-point something else (apr_memcache_server_connect_with_timeout seems good).  From a CS/SE perspective this is offensive and we admit it, but from a product quality perspective we believe this beats freezes and complicated/imperfect workarounds with threads.

So I have two questions for the Apache community:
  1. What are the practical problems with this approach?  Note that in any case a new APR rev would require editing/ifdefing our code anyway, so I think immunity from APR updates such as this patch being applied is not a distinguishing drawback.
  2. Is there an example of the correct solution to the technical problem Jeff highlighted: "it is apparently missing a call to adjust the socket timeout and to discard the connection if the timeout is reached".  That sounds like a pattern that might be found elsewhere in the Apache HTTPD code base.
Thanks in advance for your help!

On Wed, Oct 17, 2012 at 8:16 PM, Jeff Trawick <> wrote:
On Wed, Oct 17, 2012 at 3:36 PM, Joshua Marantz <> wrote:
> Is there a mechanism to time out individual operations?

No, the socket connect timeout is hard-coded at 1 second and the
socket I/O timeout is disabled.

Bugzilla bug
has a patch, though it is apparently missing a call to adjust the
socket timeout and to discard the connection if the timeout is
reached.  More importantly, the API can only be changed in future APR
2.0; alternate, backwards-compatible API(s) could be added in future
APR-Util 1.6.

> If memcached freezes, then it appears my calls to 'get' will block until
> memcached wakes up.  Is there any way to set a timeout for that call?
> I can repro this in my unit tests by sending a SIGSTOP to memcached before
> doing a 'get'?
> Here are my observations:
> apr_memcache_multgetp seems to time out in bounded time if I SIGSTOP the
> memcached process. Yes!
> apr_memcache_getp seems to hang indefinitely if I SIGSTOP the memcached
> process.
> apr_memcache_set seems to hang indefinitely if I SIGSTOP the memcached
> process.
> apr_memcache_delete seems to hang indefinitely if I SIGSTOP the memcached
> process.
> apr_memcache_stats seems to hang indefinitely if I SIGSTOP the memcached
> process.
> That last one really sucks as I am using that to print the status of all my
> cache shards to the log file if I detected a problem :(
> Why does apr_memcache_multgetp do what I want and not the others?  Can I
> induce the others to have reasonable timeout behavior?
> When I SIGSTOP memcached this makes it hard to even restart Apache, at
> least with graceful-stop.
> On a related note, the apr_memcache
> documentation<>is
> very thin.  I'd be happy to augment it with my observations on its
> usage
> and the meaning of some of the arguments if that was desired.  How would I
> go about that?

Check out APR trunk from Subversion, adjust the doxygen docs in the
include files, build (make dox) and inspect the results, submit a
patch to

> -Josh

Born in Roswell... married an alien...