Re: Persistence in effect only in case of a crash?

[Keith Rarick]

On Tue, Nov 20, 2012 at 1:49 PM, Ben <googl...@niessen.ch> wrote:
> I started beanstalkd using "-b" to have a binlog created. Now I expected
> that every job which is not yet finished/deleted will be restored after a
> restart of beanstalkd. In fact the jobs will only be recovered if I kill
> beanstalkd via "killall beanstalkd" instead of a clean shutdown.

I'm not sure I understand what you mean by "clean shutdown".
Beanstalkd doesn't distinguish different ways of shutting it down;
it's crash-only.

Can you provide specific steps reproduce the behavior you don't expect?

If you expected "every job which is not yet finished/deleted will be restored",
can you describe what you actually saw? (For example, "after starting,
beanstalkd was empty, with no jobs counted in the stats command
output".)

[Ben]

I am running FreeBSD 9 and the following happens:

1. I create a job in a queue called "test".

2. I run /usr/local/etc/rc.d/beanstalkd stop

3. I run /usr/local/etc/rc.d/beanstalkd start

4. The job is gone (no worker who could process the task is online)

Now I do the same but change the way I stop beanstalkd:

1. I create a job in a queue called "test".

2. I run killall beanstalkd

3. I run /usr/local/etc/rc.d/beanstalkd start

4. The job is still in the queue and can be processed

So for me this is a different behavior. I called the stop script "clean".

Why is this different?

Thanks for your reply.

Am Dienstag, 20. November 2012 22:49:00 UTC+1 schrieb Ben:

Hi,

I started beanstalkd using "-b" to have a binlog created. Now I expected that every job which is not yet finished/deleted will be restored after a restart of beanstalkd. In fact the jobs will only be recovered if I kill beanstalkd via "killall beanstalkd" instead of a clean shutdown.

Is this the expected behavior? Wouldn't it make sense to keep the log in any case?

I am running Beanstalkd 1.6.

Thanks for help.

[Chad Kouse]

What's in your beanstalkd init script? 

-- 
Chad Kouse

Sent with Sparrow

--
You received this message because you are subscribed to the Google Groups "beanstalk-talk" group.
To view this discussion on the web visit https://groups.google.com/d/msg/beanstalk-talk/-/_T4CVoKhzLMJ.
To post to this group, send email to beansta...@googlegroups.com.
To unsubscribe from this group, send email to beanstalk-tal...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/beanstalk-talk?hl=en.

[Ben]

I don't think it's much of a help as the actual stop command is not present (only poststop):

#!/bin/sh
#
# $FreeBSD: ports/net/beanstalkd/files/beanstalkd.in,v 1.4 2012/02/17 16:49:41 vsevolod Exp $
#
# PROVIDE: beanstalkd
# REQUIRE: DAEMON
# KEYWORD: shutdown
#
# Add the following lines to /etc/rc.conf to run beanstalkd:
#
# beanstalkd_enable (bool): Set it to "YES" to enable beanstalkd.
# Default is "NO".
# beanstalkd_flags (flags): Set extra flags here. More options in beanstalkd(1)
# Default is empty "".
# beanstalkd_user (user): Set user to run beanstalkd.
# Default is "nobody".
# beanstalkd_profiles (str):     Set to "" by default.
#                               Define your profiles here.
. /etc/rc.subr
name="beanstalkd"
rcvar=beanstalkd_enable
_piddir="/var/run/beanstalkd"
pidfile="${_piddir}/beanstalkd.pid"
load_rc_config ${name}
if [ -n "$2" ]; then
profile="$2"
if [ -n "${beanstalkd_profiles}" ]; then
pidfile="${_piddir}/beanstalkd.${profile}.pid"
eval beanstalkd_enable="\${beanstalkd_${profile}_enable:-${beanstalkd_enable}}"
eval beanstalkd_flags="\${beanstalkd_${profile}_flags:-${beanstalkd_flags}}"
else
echo "/usr/local/etc/rc.d/beanstalkd: extra argument ignored"
fi
else
if [ -n "${beanstalkd_profiles}" -a -n "$1" ]; then
for profile in ${beanstalkd_profiles}; do
eval _enable="\${beanstalkd_${profile}_enable}"
case "${_enable:-${beanstalkd_enable}}" in
[Nn][Oo]|[Ff][Aa][Ll][Ss][Ee]|[Oo][Ff][Ff]|0)
continue
;;
[Yy][Ee][Ss]|[Tt][Rr][Uu][Ee]|[Oo][Nn]|1)
;;
*)
if test -z "$_enable"; then
_var=beanstalkd_enable
else
_var=beanstalkd_"${profile}"_enable
fi
warn "Bad value" \
   "'${_enable:-${beanstalkd_enable}}'" \
   "for ${_var}. " \
   "Profile ${profile} skipped."
continue
;;
esac
echo "===> beanstalkd profile: ${profile}"
if /usr/local/etc/rc.d/beanstalkd $1 ${profile} ; then
success="${profile} ${success:-}"
else
failed="${profile} (${retcode}) ${failed:-}"
fi
done
exit 0
fi
fi
beanstalkd_poststop()
{
if [ -n "${profile}" ]; then
[ -e "$pidfile" ] && unlink $pidfile
else
local file
for file in ${_piddir}/* ; do
case "$file" in
*\*)
continue ;;
esac
unlink $file
done
fi
}
beanstalkd_prestart()
{
install -d -o $beanstalkd_user -g $beanstalkd_user -m 755 $_piddir
}
: ${beanstalkd_enable="NO"}
: ${beanstalkd_user="nobody"}
procname=/usr/local/bin/beanstalkd
command="/usr/sbin/daemon"
command_args="-p ${pidfile} ${procname} ${beanstalkd_flags}"
unset beanstalkd_flags
start_precmd="${name}_prestart"
stop_postcmd="${name}_poststop"
run_rc_command "$1"

I will test a few signals to find out which one the script sends to beanstalkd.

Thanks.