This is beanstalk-talk, the mailing list for developer and user discussion for beanstalkd and its client libraries. ()

Keith Rarick | 1 Oct 2009 10:24

Picon

Re: State of persistence in Beanstalkd


Hi everyone,

First, let me apologize for this taking so long. After the commits I
just pushed today, I am happy with the state of the persistence code
and beanstalkd in general and am just about ready to release 1.4.

It'd be really awesome if those using persistence could check out the
head of master from github.com/kr/beanstalkd and give it a spin. And
anyone else who wants to try out a "release candidate", please go
ahead.

I'll wait a couple of days in case anything crops up, and then go
ahead and tag the release, put it up on the web site, and post an
announcement.

Everyone who has contributed to this release and in the past, I can't
thank you enough. This bears repeating in the release announcement for
1.4, but I just wanted to express my gratitude now. This release is
going to be great. It contains a lot of fixes, a huge new feature, and
is our most portable beanstalkd so far. (And it's entirely my fault
that it has taken so long. Again sorry.) Thanks for sticking with me!

Up next... work on 1.5 with named jobs (for deduplication etc),
mailboxes (erlang-style, for job completion status etc), and maybe
other fun features folks come up with. Oh, and I'll finally get to
work on the web site once 1.4 is behind us. Hooray!

kr

(Continue reading)

igrigorik | 1 Oct 2009 14:59

Re: State of persistence in Beanstalkd


Woot! Thanks for the hard work Keith.

Really looking forward to named jobs. 

ig

On Oct 1, 4:24 am, Keith Rarick <k...@...> wrote:
> Hi everyone,
>
> First, let me apologize for this taking so long. After the commits I
> just pushed today, I am happy with the state of the persistence code
> and beanstalkd in general and am just about ready to release 1.4.
>
> It'd be really awesome if those using persistence could check out the
> head of master from github.com/kr/beanstalkd and give it a spin. And
> anyone else who wants to try out a "release candidate", please go
> ahead.
>
> I'll wait a couple of days in case anything crops up, and then go
> ahead and tag the release, put it up on the web site, and post an
> announcement.
>
> Everyone who has contributed to this release and in the past, I can't
> thank you enough. This bears repeating in the release announcement for
> 1.4, but I just wanted to express my gratitude now. This release is
> going to be great. It contains a lot of fixes, a huge new feature, and
> is our most portable beanstalkd so far. (And it's entirely my fault
> that it has taken so long. Again sorry.) Thanks for sticking with me!
>

(Continue reading)

Graham Barr | 1 Oct 2009 15:33

Picon

Re: State of persistence in Beanstalkd


On Thu, Oct 1, 2009 at 3:24 AM, Keith Rarick <kr@...> wrote:
> First, let me apologize for this taking so long. After the commits I
> just pushed today, I am happy with the state of the persistence code
> and beanstalkd in general and am just about ready to release 1.4.

There is probably a better way then hard coding in the file, but please update
the version number in beanstalkd.spec

Graham.

--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups "beanstalk-talk" group.
To post to this group, send email to beanstalk-talk@...
To unsubscribe from this group, send email to beanstalk-talk+unsubscribe@...
For more options, visit this group at http://groups.google.com/group/beanstalk-talk?hl=en
-~----------~----~----~----~------~----~------~--~---

Keith Rarick | 2 Oct 2009 02:43

Picon

Re: State of persistence in Beanstalkd

On Thu, Oct 1, 2009 at 6:33 AM, Graham Barr <gmbarr@...> wrote:
> There is probably a better way then hard coding in the file, but please update
> the version number in beanstalkd.spec

Thanks for reminding me! It's now an autoconf template and included in
the distributed tarball.

I don't know the usual flow for generating rpm files. Hopefully this
will work. I only know of one problem with this approach: the version
string between releases has hyphens in it (like 1.3-55-g9f71603
instead of just 1.3) and rpm forbids this (according to the
documentation). I welcome any suggestions on how to deal with this.

kr

--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups "beanstalk-talk" group.
To post to this group, send email to beanstalk-talk@...
To unsubscribe from this group, send email to beanstalk-talk+unsubscribe@...
For more options, visit this group at http://groups.google.com/group/beanstalk-talk?hl=en
-~----------~----~----~----~------~----~------~--~---

Jeremy Hinegardner | 2 Oct 2009 17:13

Re: State of persistence in Beanstalkd


On Thu, Oct 01, 2009 at 05:43:58PM -0700, Keith Rarick wrote:
> 
> On Thu, Oct 1, 2009 at 6:33 AM, Graham Barr <gmbarr@...> wrote:
> > There is probably a better way then hard coding in the file, but please update
> > the version number in beanstalkd.spec
> 
> Thanks for reminding me! It's now an autoconf template and included in
> the distributed tarball.
> 
> I don't know the usual flow for generating rpm files. Hopefully this
> will work. I only know of one problem with this approach: the version
> string between releases has hyphens in it (like 1.3-55-g9f71603
> instead of just 1.3) and rpm forbids this (according to the
> documentation). I welcome any suggestions on how to deal with this.

We've been using 'git describe' to populate the release number in our 
projects stored in git.  That '55' above, is the number of commits since the 
v1.3 tag.

  jeremy <at> stiletto:~/repos/git/beanstalkd master % git describe --match v1.1
  v1.1-171-g9f71603
  jeremy <at> stiletto:~/repos/git/beanstalkd master % git describe --match v1.2
  v1.2-60-g9f71603
  jeremy <at> stiletto:~/repos/git/beanstalkd master % git describe --match v1.3
  v1.3-55-g9f71603 

So from beanstalkd's perspective, you can use the 1.3 for the  <at> VERSION <at>  and
then make:

(Continue reading)

Jeremy Dunck | 2 Oct 2009 18:25

Picon

Re: State of persistence in Beanstalkd


On Fri, Oct 2, 2009 at 10:13 AM, Jeremy Hinegardner
<jeremy@...> wrote:
...
>  jeremy <at> stiletto:~/repos/git/beanstalkd master % git describe --match v1.1
>  v1.1-171-g9f71603
>  jeremy <at> stiletto:~/repos/git/beanstalkd master % git describe --match v1.2
>  v1.2-60-g9f71603
>  jeremy <at> stiletto:~/repos/git/beanstalkd master % git describe --match v1.3
>  v1.3-55-g9f71603
>
...
> Its not perfect, if you have the same number of commits from a tag in two
> different branches, you could have 2 spec files that had the same
> name-version-release but would be two different rpms.

Isn't that hash prefix at the end (g9f71603) the current commit?  It's
the same in your example because you were trying to demonstrate
git-describe, but in real life, the output of the same git-describe on
two different commits, even with the same # of commits to the same
ancestor tag, would have different hashes.

Right?

--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups "beanstalk-talk" group.
To post to this group, send email to beanstalk-talk@...
To unsubscribe from this group, send email to beanstalk-talk+unsubscribe <at> googlegroups.com
For more options, visit this group at http://groups.google.com/group/beanstalk-talk?hl=en
-~----------~----~----~----~------~----~------~--~---

(Continue reading)

Keith Rarick | 2 Oct 2009 22:32

Picon

Re: State of persistence in Beanstalkd


On Fri, Oct 2, 2009 at 8:13 AM, Jeremy Hinegardner
<jeremy@...> wrote:
> So from beanstalkd's perspective, you can use the 1.3 for the  <at> VERSION <at>  and
> then make:
>
>  Release:   <at> RELEASE <at> %{?dist}
>
> and populate  <at> RELEASE <at>  based upon the output of 'git describe'.

That is pretty clever, but I suspect that it will cause confusion.

I'm leaning toward replacing the hyphens with pluses, to get 1.3+55+g9f71603.

kr

--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups "beanstalk-talk" group.
To post to this group, send email to beanstalk-talk@...
To unsubscribe from this group, send email to beanstalk-talk+unsubscribe <at> googlegroups.com
For more options, visit this group at http://groups.google.com/group/beanstalk-talk?hl=en
-~----------~----~----~----~------~----~------~--~---

Dustin | 2 Oct 2009 23:50

Picon

Re: State of persistence in Beanstalkd


On Oct 2, 1:32 pm, Keith Rarick <k...@...> wrote:

> I'm leaning toward replacing the hyphens with pluses, to get 1.3+55+g9f71603.

  In memcached, I use underscore, but it's slightly inconvenient
because I have to rewrite them all to hyphens again.  That's an
amazingly easy fix in git, though.  I should really do that...
--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups "beanstalk-talk" group.
To post to this group, send email to beanstalk-talk@...
To unsubscribe from this group, send email to beanstalk-talk+unsubscribe <at> googlegroups.com
For more options, visit this group at http://groups.google.com/group/beanstalk-talk?hl=en
-~----------~----~----~----~------~----~------~--~---

Keith Rarick | 2 Oct 2009 23:51

Picon

Re: TTR is not enforced if socket is not active?


By the way, hi res timing is in commit
http://github.com/kr/beanstalkd/commit/6e9339ff9c177cc37787ee252fd3cf3a04f17243
.

This should fix all the timeout problems people have seen.

kr

On Wed, Jun 3, 2009 at 8:58 AM, Keith Rarick <kr@...> wrote:
> Yes, I think your analysis is correct.
>
> On Wed, Jun 3, 2009 at 8:24 AM, Yun Huang Yong <yun@...> wrote:
>> The only real fix for this is to use higher resolution
>> timings as noted in
>> http://github.com/kr/beanstalkd/issues#issue/5.
>>
>> Keith, how far away is the higher resolution timeout stuff? :)
>
> Not too far, but I'd like to get the binlog code stable first, unless
> this is causing serious problems.
>
> Sorry things have been so quiet lately; I've been hired to work on
> another project and that has kept me busy. (But it's a pretty cool
> project. I'll be announcing it soon.)
>
> kr
>

--~--~---------~--~----~------------~-------~--~----~

(Continue reading)

Keith Rarick | 3 Oct 2009 14:04

Picon

[ANN] beanstalkd 1.4


Okay, beanstalkd 1.4 is ready.

To everyone who has contributed to this release and in the past, I can't thank
you enough. This release contains a lot of fixes, a huge new feature, and is
our most portable beanstalkd so far. Thanks for sticking with me!

There will be no incompatible protocol changes until version 2.0. A client
written for version 1.4 will work unmodified with any later 1.x release of
beanstalkd.

What's New
----------

 * Persistence (Graham Barr, Keith Rarick)
 * Properly initialize network address struct (Jeethu Rao)
 * Fix daemonization on BSD systems, including Mac OS X (Alex MacCaw and Yun
   Huang Yong)
 * Now includes RPM spec file and init scripts (Ask Bjørn Hansen)
 * Fixed build and unit tests on Solaris and Mac (Keith Rarick)
 * Added "reserves" counter for each job (goosmurf)
 * Fix a memory leak from the peek and stats commands (goosmurf)
 * Handle SIGTERM just like SIGINT (Florian Noeding)
 * Use high-resolution timers, fixing some timeout-related bugs (Keith Rarick)

Our Urls
--------

Download the 1.4 tarball directly:
<http://xph.us/dist/beanstalkd/beanstalkd-1.4.tar.gz>

(Continue reading)

Group

gmane.network.beanstalk.general.

Project Web Page

This is beanstalk-talk, the mailing list for developer and user discussion for beanstalkd and its client libraries. ()

Search Archive

Page

1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10

Articles in period: 95

October 2009

Language

Change language

Options

Current view: Threads only / Showing only 20 lines / Not hiding cited text.
Change to All messages, whole messages, or hide cited text.

Post a message
NNTP Newsgroup
Classic Gmane web interface
XML

XML

RSS Feed
List Information

About Gmane

Recent entries

Beantop: A command line tool (23 May 2013)
Can I use beanstalkd to queue up tasks that can be executed at a speci (20 May 2013)
How many queues can beanstalkd handle? (13 May 2013)
"touch" command (29 Apr 2013)
[ANN] beanstalkd 1.9 (14 Apr 2013)
Version 1.9 (12 Apr 2013)
Deleting binlog files (1 Apr 2013)
Stateless put and reserve commands (30 Mar 2013)
how to list out all the workers and producers for a tube (22 Mar 2013)
Get all job ids in specific tube (21 Mar 2013)
Running beanstalkd in init script with persistence option. (10 Mar 2013)
Is version 1.8 stable? (12 Feb 2013)
Beanstalkd uses a lot more CPU cycles when the system is under high lo (12 Feb 2013)
beanstalkd writes 100+ binlogs then stops responding (9 Feb 2013)
Slow unless on loopback (17 Dec 2012)
core dump with beanstalkd 1.8 (13 Dec 2012)
Routeing Jobs (5 Dec 2012)
General approach for multiple worker types (5 Dec 2012)
Persistence in effect only in case of a crash? (20 Nov 2012)
beanstemc.reserve(60) works but beantstemc.reserve() doesnt work as ex (15 Nov 2012)
who is using beanstalkd in production? (10 Nov 2012)

Archives

10	May 2013
18	April 2013
8	March 2013
6	February 2013
6	January 2013
20	December 2012
30	November 2012
53	October 2012
12	September 2012
15	August 2012
14	July 2012
5	June 2012
14	May 2012
22	April 2012
10	March 2012
43	February 2012
31	January 2012
20	December 2011
16	November 2011
5	October 2011
23	September 2011
30	August 2011
18	July 2011
16	June 2011
8	May 2011
6	April 2011
10	March 2011
27	February 2011
29	January 2011
9	December 2010
59	November 2010
12	October 2010
24	September 2010
56	August 2010
32	July 2010
34	June 2010
26	May 2010
14	April 2010
63	March 2010
25	February 2010
27	January 2010
72	December 2009
113	November 2009
95	October 2009
44	September 2009
59	August 2009
14	July 2009
13	June 2009
24	May 2009
41	April 2009
8	March 2009
27	February 2009
69	January 2009
29	December 2008
14	November 2008

Design

Zawodny | Right Menu | Left Menu

Your Own Design

Paste the URL of your CSS below. Download CSS template