INTERNAL_ERROR

2 views
Skip to first unread message

Erwan A.

unread,
Nov 17, 2009, 2:14:48 PM11/17/09
to beanstalk-talk
Hello all,

i have just compiled a fresh beanstalkd from github on OSX and i'm
getting a consistent INTERNAL_ERROR when running the following test
(in ruby)

def insert_a_few_jobs(bs,number)
bs.use 'test'
number.times do |i|
body = "test job number #{i}"
bs.put body
end
bs.watch 'test'
number.times do |i|
body = "test buried job number #{i}"
bs.put body
job = bs.reserve 0
job.bury #### crashes here.
end
end

Stacktrace (client side)

test_export(TestBeanstalkServer):
Beanstalk::InternalError: INTERNAL_ERROR
method check_resp in connection.rb at line 197
method interact in connection.rb at line 184
method bury in connection.rb at line 116
method bury in job.rb at line 64
method insert_a_few_jobs in tc_beanstalk_server.rb at line 50
method times in tc_beanstalk_server.rb at line 46
method insert_a_few_jobs in tc_beanstalk_server.rb at line 46
method test_export in tc_beanstalk_server.rb at line 57
method test_export in tc_beanstalk_server.rb at line 55


and i get the following output on beanstalkd's stderr.

/opt/beanstalkd-1.4.2/bin/beanstalkd: binlog.c:457 in
binlog_write_job: unserializable job state: 3
/opt/beanstalkd-1.4.2/bin/beanstalkd: prot.c:1292 in dispatch_cmd:
server error: INTERNAL_ERROR

beanstalkd command line is :
/opt/beanstalkd-1.4.2/bin/beanstalkd -b /var/folders/0z/
0zkMgRtCGm8ofh4LX0jh0U+++TI/-Tmp-/beanstalk-
logbin20091117-15371-1d6o2wwtmp -l 127.0.0.1 -p 45713 -f 100

(it's a local server that gets killed as soon as the unit test is
complete, thus the strange location of the binlog file)

I'm currently evaluating upgrading from 1.3, am writing the necessary
import/export tools, and i need to be able to bury a job as soon as it
is pushed in order to recover our active queues in exactly the same
state.

Anything i'm doing wrong ?

Thanks in advance,

Erwan

Erwan A.

unread,
Nov 18, 2009, 4:52:20 AM11/18/09
to beanstalk-talk
On 17 nov, 20:14, "Erwan A." <ear...@gmail.com> wrote:
> Hello all,
>
> i have just compiled a fresh beanstalkd from github on OSX and i'm
> getting a consistent INTERNAL_ERROR when running the following test
> (in ruby)
>
....
>
> Anything i'm doing wrong ?
>

Could this be caused by the ruby client ? Should we evaluate switching
to em-jack as Keith suggested in another message ?

Thanks,

Erwan

Keith Rarick

unread,
Nov 28, 2009, 6:08:50 AM11/28/09
to beansta...@googlegroups.com
On Tue, Nov 17, 2009 at 11:14 AM, Erwan A. <ear...@gmail.com> wrote:
> Anything i'm doing wrong ?

No, this is a bug in beanstalkd. I'm fixing it now, and the fix will
be in the next release, which should be out in a day or two, unless
other bugs pop up.

kr
Reply all
Reply to author
Forward
0 new messages