You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@serf.apache.org by "Simon Bin (JIRA)" <ji...@apache.org> on 2016/11/29 10:38:58 UTC
[jira] [Issue Comment Deleted] (SERF-166) subversion + serf +
kerberos + https hangs
[ https://issues.apache.org/jira/browse/SERF-166?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Simon Bin updated SERF-166:
---------------------------
Comment: was deleted
(was: Here is one backtrace of such a hang
#0 0x00007fbad3c1ef1c in __lll_lock_wait () at ../sysdeps/unix/sysv/linux/x86_64/lowlevellock.S:135
#1 0x00007fbad3c18b45 in __GI___pthread_mutex_lock (mutex=0x564e97996000) at ../nptl/pthread_mutex_lock.c:80
__PRETTY_FUNCTION__ = "__pthread_mutex_lock"
type = 0
id = <optimized out>
#2 0x00007fbad3e49099 in apr_allocator_alloc () at /usr/lib64/libapr-1.so.0
#3 0x00007fbad15c393c in serf_bucket_mem_alloc () at /usr/lib64/libserf-1.so.1
#4 0x00007fbad15c7b17 in serf_bucket_response_create () at /usr/lib64/libserf-1.so.1
#5 0x00007fbad15c2217 in serf.process_connection () at /usr/lib64/libserf-1.so.1
#6 0x00007fbad15c068e in serf_event_trigger () at /usr/lib64/libserf-1.so.1
#7 0x00007fbad15c07c9 in serf_context_run () at /usr/lib64/libserf-1.so.1
#8 0x00007fbad3008be1 in svn_ra_serf__context_run (sess=sess@entry=0x564e975fadc0, waittime_left=waittime_left@entry=0x7ffd03fee9a8, scratch_pool=scratch_pool@entry=0x564e97658528) at subversion/libsvn_ra_serf/util.c:866
status = <optimized out>
err = <optimized out>
scratch_pool = 0x564e97658528
waittime_left = 0x7ffd03fee9a8
sess = 0x564e975fadc0
#9 0x00007fbad30074b8 in finish_report (scratch_pool=0x564e97610238, handler=0x564e976103d0, ctx=0x564e975e9008) at subversion/libsvn_ra_serf/update.c:2614
err = <optimized out>
i = <optimized out>
sess = 0x564e975fadc0
iterpool = 0x564e97658528
waittime_left = 600000000
report = 0x564e975e9008
sess = <optimized out>
handler = 0x564e976103d0
xmlctx = <optimized out>
report_target = 0x564e976102f8 "/projects/!svn/me"
buf = 0x564e976102b0
scratch_pool = 0x564e97610238
err = <optimized out>
#10 0x00007fbad30074b8 in finish_report (report_baton=0x564e975e9008, pool=<optimized out>) at subversion/libsvn_ra_serf/update.c:2688
report = 0x564e975e9008
sess = <optimized out>
handler = 0x564e976103d0
xmlctx = <optimized out>
report_target = 0x564e976102f8 "/projects/!svn/me"
buf = 0x564e976102b0
scratch_pool = 0x564e97610238
err = <optimized out>
)
> subversion + serf + kerberos + https hangs
> ------------------------------------------
>
> Key: SERF-166
> URL: https://issues.apache.org/jira/browse/SERF-166
> Project: serf
> Issue Type: Bug
> Reporter: Serf Importer
> Labels: Priority-Medium, Type-Defect
>
> Checking out a medium sized repo using kerberos and https
> hangs in mid air. Kill with Ctrl-C and do svn update and it too
> hangs before finished.
> However, following this advise about apache:
> The MaxKeepAliveRequests option in httpd.conf needs to be increased from 100 (the default) to at least 1000 (there is no reason why it could not be 10000). This will improve performance by allowing serf clients to use fewer TCP connections to the server. Clients using neon will also work fine with this configuration.
> Makes the hangs go away. Seems like there is a bug in serf handling low numbers
> for MaxKeepAliveRequests
> subversion 1.8.13 and serf 1.3.8
> Original issue reported by *joakim.tjernlund*
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)