Re: Slony-I 1.0.4 Released

The Slony-I team is proud to present the 1.0.4 release of the most
advanced replication solution for the most advanced Open Source
Database in the world.

The release tarball is available for download
http://developer.postgresql.org/~wieck/slony1/download/slony1-1.0.4.tar.gz

There are a limited number of "new features" this release largely in
terms of adding in the ability to move or drop tables and sequences
from replication:

- SET DROP TABLE - drops a table from replication
- SET DROP SEQUENCE - does the same for sequences
- SET MOVE TABLE - moves a table from one replication set to another
- SET MOVE SEQUENCE - moves a sequence from one replication set to another

Other changes involve smoothing out the 'sharp edges' found by early
adopters, notably including:

- Frequently vacuuming pg_listener; growth of dead tuples could hurt
performance

- A cleanup process for pg_listener resolves cases where old slon
processes may have terminated due to network problems, leaving
backends around holding onto event notifications

- Lowered lock level on sl_event, resolving issues where pg_dump
would block Slony-I

- Purges CONFIRM entries for nodes that don't exist anymore

- Substantially increased documentation

- More sophisticated administration scripts

- Now uses string comparison for user defined types that do not have
a suitable comparison operation

- Safer log purging

- Various other bug fixes and "improved cleanliness."

- As of 1.0.4, the slon replication engine refuses to work against
any database that does not have the stored procedures for the same
version loaded or where the shared object containing the C language
support functions and the log trigger does not match the version
number. in a cluster must be upgraded at once.

See the HISTORY-1.0 file for a detailed list of changes.

<http://slony.info/>
--
let name="cbbrowne" and tld="cbbrowne.com" in String.concat "@" [name;tld];;
http://www.ntlug.org/~cbbrowne/linuxxian.html
A VAX is virtually a computer, but not quite.

Chris Browne wrote:
> The Slony-I team is proud to present the 1.0.4 release of the most
> advanced replication solution for the most advanced Open Source
> Database in the world.
>
> The release tarball is available for download
> http://developer.postgresql.org/~wieck/slony1/download/slony1-1.0.4.tar.gz

May I use it on a machine with a postgres 7.4.5 installed with RPM ?

Regards
Gaetano Mendola

Oops! Gaetano Mendola <mendola(at)bigfoot(dot)com> was seen spray-painting on a wall:
> Chris Browne wrote:
> > The Slony-I team is proud to present the 1.0.4 release of the most
> > advanced replication solution for the most advanced Open Source
> > Database in the world.
> >
> > The release tarball is available for download
> > http://developer.postgresql.org/~wieck/slony1/download/slony1-1.0.4.tar.gz
>
> May I use it on a machine with a postgres 7.4.5 installed with RPM ?

Probably only with some degree of trickery.

In the "1.0" series, it needs a copy of the PostgreSQL source tree. I
believe it uses that to refer it to where its components should be
installed.

In 1.1 (in CVS HEAD), it takes some advantage of the "pgxs" support,
so that you may instead point to the #includes, libs, and binaries.
But that's definitely a "work in progress," as opposed to something
that is complete.

All I can say for certain is that getting this sort of thing to work
on AIX is not fun, whatever the approach :-(.
--
output = reverse("gro.gultn" "@" "enworbbc")
http://www.ntlug.org/~cbbrowne/sap.html
Rules of the Evil Overlord #43. "I will maintain a healthy amount of
skepticism when I capture the beautiful rebel and she claims she is
attracted to my power and good looks and will gladly betray her
companions if I just let her in on my plans."
<http://www.eviloverlord.com/>

Sorry folks,

the Slony-I team has produced a great product, but the project
management (that's mostly me here) sucks big time!

Shortly after giving Chris Browne green light for the 1.0.4 announcement
we found a way to guard against bug #896. That being a really bad one I
decided to stop the 1.0.4 release and go for 1.0.5 including that fix.
But I failed to make sure Chris, Justin and others involved in the
announcement process get the message.

I have just committed those changes and 1.0.5 should be available later
today.

I apologize for the confusion and look forward to retire from the
Slony-I project leadership position in order to fully focus on the
multimaster replication project Afilias has decided to put forward.

Jan

On 10/21/2004 5:55 PM, Chris Browne wrote:

> The Slony-I team is proud to present the 1.0.4 release of the most
> advanced replication solution for the most advanced Open Source
> Database in the world.
>
> The release tarball is available for download
> http://developer.postgresql.org/~wieck/slony1/download/slony1-1.0.4.tar.gz
>
> There are a limited number of "new features" this release largely in
> terms of adding in the ability to move or drop tables and sequences
> from replication:
>
> - SET DROP TABLE - drops a table from replication
> - SET DROP SEQUENCE - does the same for sequences
> - SET MOVE TABLE - moves a table from one replication set to another
> - SET MOVE SEQUENCE - moves a sequence from one replication set to another
>
> Other changes involve smoothing out the 'sharp edges' found by early
> adopters, notably including:
>
> - Frequently vacuuming pg_listener; growth of dead tuples could hurt
> performance
>
> - A cleanup process for pg_listener resolves cases where old slon
> processes may have terminated due to network problems, leaving
> backends around holding onto event notifications
>
> - Lowered lock level on sl_event, resolving issues where pg_dump
> would block Slony-I
>
> - Purges CONFIRM entries for nodes that don't exist anymore
>
> - Substantially increased documentation
>
> - More sophisticated administration scripts
>
> - Now uses string comparison for user defined types that do not have
> a suitable comparison operation
>
> - Safer log purging
>
> - Various other bug fixes and "improved cleanliness."
>
> - As of 1.0.4, the slon replication engine refuses to work against
> any database that does not have the stored procedures for the same
> version loaded or where the shared object containing the C language
> support functions and the log trigger does not match the version
> number. in a cluster must be upgraded at once.
>
> See the HISTORY-1.0 file for a detailed list of changes.
>
> <http://slony.info/>

--
#======================================================================#
# It's easier to get forgiveness for being wrong than for being right. #
# Let's break this rule - forgive me. #
#================================================== JanWieck(at)Yahoo(dot)com #

On 10/22/2004 11:29 AM, Ed L. wrote:

> Wow. First, thanks again for all your efforts, Jan. Second, I'm
> disappointed to hear the slony author and lead developer is leaving the
> slony leadership. When is that going to happen? And what does that mean
> with respect to your future involvement in slony?

It means that I will try to do with Slony-I what I have done
successfully with several other PostgreSQL related projects - to make
sure the project doesn't depend on me in person any more. None of the
big items I've done (rewrite rule system, PL/pgSQL, PL/Tcl, TOAST,
NUMERIC, foreign keys) really requires my attention if there's a bug. I
was able to do the conceptual work and important parts of the
implementation, but I didn't get stuck in/with the maintenance. I
constider this a very important aspect of being a CORE developer.

What I am trying to do is to spend less and less time on Slony-I and
more and more time on Slony-II - the synchronous multi-master system.
Slony-I was for sure one of the better things that I've done so far, but
it's not going to allow early retirement with financial independance and
wealth. And as long as Afilias is using Slony-I in production, Andrew
Sullivan will not let me do whatever I want if there's a severe problem
nobody else can fix.

So don't worry, I'll be around.

Jan

>
> Ed
>
>
> On Friday October 22 2004 7:26, Jan Wieck wrote:
>> Sorry folks,
>>
>> the Slony-I team has produced a great product, but the project
>> management (that's mostly me here) sucks big time!
>>
>> Shortly after giving Chris Browne green light for the 1.0.4 announcement
>> we found a way to guard against bug #896. That being a really bad one I
>> decided to stop the 1.0.4 release and go for 1.0.5 including that fix.
>> But I failed to make sure Chris, Justin and others involved in the
>> announcement process get the message.
>>
>> I have just committed those changes and 1.0.5 should be available later
>> today.
>>
>> I apologize for the confusion and look forward to retire from the
>> Slony-I project leadership position in order to fully focus on the
>> multimaster replication project Afilias has decided to put forward.
>>
>>
>> Jan
>>
>> On 10/21/2004 5:55 PM, Chris Browne wrote:
>> > The Slony-I team is proud to present the 1.0.4 release of the most
>> > advanced replication solution for the most advanced Open Source
>> > Database in the world.
>> >
>> > The release tarball is available for download
>> >
>> > http://developer.postgresql.org/~wieck/slony1/download/slony1-1.0.4.tar
>> >.gz
>> >
>> > There are a limited number of "new features" this release largely in
>> > terms of adding in the ability to move or drop tables and sequences
>> > from replication:
>> >
>> > - SET DROP TABLE - drops a table from replication
>> > - SET DROP SEQUENCE - does the same for sequences
>> > - SET MOVE TABLE - moves a table from one replication set to another
>> > - SET MOVE SEQUENCE - moves a sequence from one replication set to
>> > another
>> >
>> > Other changes involve smoothing out the 'sharp edges' found by early
>> > adopters, notably including:
>> >
>> > - Frequently vacuuming pg_listener; growth of dead tuples could hurt
>> > performance
>> >
>> > - A cleanup process for pg_listener resolves cases where old slon
>> > processes may have terminated due to network problems, leaving
>> > backends around holding onto event notifications
>> >
>> > - Lowered lock level on sl_event, resolving issues where pg_dump
>> > would block Slony-I
>> >
>> > - Purges CONFIRM entries for nodes that don't exist anymore
>> >
>> > - Substantially increased documentation
>> >
>> > - More sophisticated administration scripts
>> >
>> > - Now uses string comparison for user defined types that do not have
>> > a suitable comparison operation
>> >
>> > - Safer log purging
>> >
>> > - Various other bug fixes and "improved cleanliness."
>> >
>> > - As of 1.0.4, the slon replication engine refuses to work against
>> > any database that does not have the stored procedures for the same
>> > version loaded or where the shared object containing the C language
>> > support functions and the log trigger does not match the version
>> > number. in a cluster must be upgraded at once.
>> >
>> > See the HISTORY-1.0 file for a detailed list of changes.
>> >
>> > <http://slony.info/>

--
#======================================================================#
# It's easier to get forgiveness for being wrong than for being right. #
# Let's break this rule - forgive me. #
#================================================== JanWieck(at)Yahoo(dot)com #

Wow. First, thanks again for all your efforts, Jan. Second, I'm
disappointed to hear the slony author and lead developer is leaving the
slony leadership. When is that going to happen? And what does that mean
with respect to your future involvement in slony?

Ed

On Friday October 22 2004 7:26, Jan Wieck wrote:
> Sorry folks,
>
> the Slony-I team has produced a great product, but the project
> management (that's mostly me here) sucks big time!
>
> Shortly after giving Chris Browne green light for the 1.0.4 announcement
> we found a way to guard against bug #896. That being a really bad one I
> decided to stop the 1.0.4 release and go for 1.0.5 including that fix.
> But I failed to make sure Chris, Justin and others involved in the
> announcement process get the message.
>
> I have just committed those changes and 1.0.5 should be available later
> today.
>
> I apologize for the confusion and look forward to retire from the
> Slony-I project leadership position in order to fully focus on the
> multimaster replication project Afilias has decided to put forward.
>
>
> Jan
>
> On 10/21/2004 5:55 PM, Chris Browne wrote:
> > The Slony-I team is proud to present the 1.0.4 release of the most
> > advanced replication solution for the most advanced Open Source
> > Database in the world.
> >
> > The release tarball is available for download
> >
> > http://developer.postgresql.org/~wieck/slony1/download/slony1-1.0.4.tar
> >.gz
> >
> > There are a limited number of "new features" this release largely in
> > terms of adding in the ability to move or drop tables and sequences
> > from replication:
> >
> > - SET DROP TABLE - drops a table from replication
> > - SET DROP SEQUENCE - does the same for sequences
> > - SET MOVE TABLE - moves a table from one replication set to another
> > - SET MOVE SEQUENCE - moves a sequence from one replication set to
> > another
> >
> > Other changes involve smoothing out the 'sharp edges' found by early
> > adopters, notably including:
> >
> > - Frequently vacuuming pg_listener; growth of dead tuples could hurt
> > performance
> >
> > - A cleanup process for pg_listener resolves cases where old slon
> > processes may have terminated due to network problems, leaving
> > backends around holding onto event notifications
> >
> > - Lowered lock level on sl_event, resolving issues where pg_dump
> > would block Slony-I
> >
> > - Purges CONFIRM entries for nodes that don't exist anymore
> >
> > - Substantially increased documentation
> >
> > - More sophisticated administration scripts
> >
> > - Now uses string comparison for user defined types that do not have
> > a suitable comparison operation
> >
> > - Safer log purging
> >
> > - Various other bug fixes and "improved cleanliness."
> >
> > - As of 1.0.4, the slon replication engine refuses to work against
> > any database that does not have the stored procedures for the same
> > version loaded or where the shared object containing the C language
> > support functions and the log trigger does not match the version
> > number. in a cluster must be upgraded at once.
> >
> > See the HISTORY-1.0 file for a detailed list of changes.
> >
> > <http://slony.info/>

The Slony-I team <http://slony.info/> is happy to present the 1.0.5
release.

The release tarball is available for download
http://developer.postgresql.org/~wieck/slony1/download/slony1-1.0.5.tar.gz

See the HISTORY-1.0 file for a detailed list of changes.
--
let name="cbbrowne" and tld="cbbrowne.com" in String.concat "@" [name;tld];;
http://www.ntlug.org/~cbbrowne/linuxxian.html
A VAX is virtually a computer, but not quite.

On Fri, Oct 22, 2004 at 09:36:10AM -0400, Jan Wieck wrote:
> wealth. And as long as Afilias is using Slony-I in production, Andrew
> Sullivan will not let me do whatever I want if there's a severe problem
> nobody else can fix.

Or someone around here will, anyway. I might get hit by a wayward
streetcar, but Afilias is very strongly committed to maintaining
and improving Slony-I even as we also sponsor work on multimaster
systems. Afilias uses PostgreSQL for almost everything we do;
for us, there is no question that we'll need to support Slony.

It's worth pointing out here that the really big database vendors
usually have more than one thing that they sell as "replication".
That's because you can't really use (for instance) RAC across a WAN.
But if you're running a bank, or network infrastructure, or a phone
company, or anything else that requires you to have plans for the
disaster recovery of your disaster recovery site, you'll certainly
need to have replicas which are geographically dispersed. That's why
we've assigned staff, aside from Jan, to keep working on Slony-I.
That's _also_ why we wanted, from the very beginning, to build a
community around Slony: we knew that we couldn't do it ourselves, if
only because our perspective is bound to be limited by our
experiences, and we wanted to make sure that we wouldn't build in
such limitations. I think we've been pretty successful in that
attempt, and I'm delighted that we've managed to attract many
really smart people to the project.

A

--
Andrew Sullivan | ajs(at)crankycanuck(dot)ca
This work was visionary and imaginative, and goes to show that visionary
and imaginative work need not end up well.
--Dennis Ritchie

Christopher Browne wrote:
> Oops! Gaetano Mendola <mendola(at)bigfoot(dot)com> was seen spray-painting on a wall:
>
>>Chris Browne wrote:
>> > The Slony-I team is proud to present the 1.0.4 release of the most
>> > advanced replication solution for the most advanced Open Source
>> > Database in the world.
>> >
>> > The release tarball is available for download
>> > http://developer.postgresql.org/~wieck/slony1/download/slony1-1.0.4.tar.gz
>>
>>May I use it on a machine with a postgres 7.4.5 installed with RPM ?
>
>
> Probably only with some degree of trickery.

I suspected it, I currently can not use it because of this. Any chance to
have a slony rpm compatible with the 7.4.5 rpm ?

Regards
Gaetano Mendola

pgsql(at)bluepolka(dot)net ("Ed L.") wrote:
> Wow. First, thanks again for all your efforts, Jan. Second, I'm
> disappointed to hear the slony author and lead developer is leaving
> the slony leadership. When is that going to happen? And what does
> that mean with respect to your future involvement in slony?

In some respects, it's a healthy thing for there to be _some_ change
that takes place. A project that is comprised of one person can
pretty easily disappear if that one person steps away. Furthermore,
one of the purported merits of open source is that it benefits from
diverse involvement. A healthy OSS project sees changes coming in
from _multiple_ people; from the diversity of people comes a diversity
in points of view and in ideas of what can be added. That has already
been the case with Slony-I.

Furthermore, I don't see this leading to Jan "disappearing from Slony"
at all. As one who 'groks the code well,' his wisdom will doubtless
be drawn upon. (And since those paying his salary depend quite a lot
on Slony-I, there's a certain amount of influence that could be
brought to bear ;-).) He's trying to spend some time concentrating on
yet another project that is also called Slony, namely Slony-II.

The point of Slony-I was to provide an efficient way to bring a bunch
of nodes into sync. Slony-II is to extend that further to support
multimaster replication. It needs to start with databases that are in
sync, ergo requiring starting with Slony-I. Others of us may be doing
most of the "heavy lifting" to finish off features already
contemplated for Slony-I, but I seriously doubt that means Jan will be
completely uninvolved.

There needs to be an interface between the phases, too. After all,
there has to be some transformation to take a set of "Slony-I
replicated" systems to then take them into "Slony-II mode." I'll bet
that means that Slony-I will have to add some code to become
"Slony-II-aware"...
--
output = ("cbbrowne" "@" "acm.org")
http://www.ntlug.org/~cbbrowne/
"Markets can remain irrational longer than you can remain solvent."
-- J. M. Keynes

After takin a swig o' Arrakan spice grog, Gaetano Mendola <mendola(at)bigfoot(dot)com> belched out:
> Christopher Browne wrote:
> > Oops! Gaetano Mendola <mendola(at)bigfoot(dot)com> was seen spray-painting on a wall:
> >
> >>Chris Browne wrote:
> >> > The Slony-I team is proud to present the 1.0.4 release of the most
> >> > advanced replication solution for the most advanced Open Source
> >> > Database in the world.
> >> >
> >> > The release tarball is available for download
> >> > http://developer.postgresql.org/~wieck/slony1/download/slony1-1.0.4.tar.gz
> >>
> >>May I use it on a machine with a postgres 7.4.5 installed with RPM ?

> > Probably only with some degree of trickery.

> I suspected it, I currently can not use it because of this. Any
> chance to have a slony rpm compatible with the 7.4.5 rpm ?

If someone contributes RPM packages, using the same GCC and the same
RPM source RPM, then presumably it's possible.

There's a BSD port, so it has proven of enough interest to attract
contributions there. There's not yet a Debian package or an RPM for
one or another of the RPM-based distributions; hopefully interested
parties will find opportunity to build such.

I can appreciate the desire to have some pre-packaged components; I
generally despise the idea of compiling Perl stuff myself as that
tends to turn into a system management nightmare.

It may be that packages will emerge in six months; we'll see. If you
need RHAT RPM packages badly enough, that may be reason to pay someone
to build them.
--
let name="cbbrowne" and tld="linuxfinances.info" in String.concat "@" [name;tld];;
http://www.ntlug.org/~cbbrowne/linuxxian.html
Y2K conversion simplified: Januark, Februark, March, April, Mak, June,
Julk, August, September, October, November, December.

On Fri, 2004-10-22 at 22:26 -0400, Christopher Browne wrote:
> > I suspected it, I currently can not use it because of this. Any
> > chance to have a slony rpm compatible with the 7.4.5 rpm ?
>
> If someone contributes RPM packages, using the same GCC and the same
> RPM source RPM, then presumably it's possible.
>
> There's a BSD port, so it has proven of enough interest to attract
> contributions there. There's not yet a Debian package or an RPM for
> one or another of the RPM-based distributions; hopefully interested
> parties will find opportunity to build such.

I'm interested in providing a Debian package, but the need to have the
postgresql source tree available is a problem. At the moment, unless I
incorporate slony-I into the postgresql source package, I can't package
it; the Debian autobuilders cannot be configured to build one package
(postgresql) and then keep its source around while another package is
built.

I thought someone was working on making a subsection of postgresql that
would be sufficient to let packages like Slony-I build independently; is
that making any progress? If not, what files from the source tree does
Slony-I need for its build? I could make a binary package to provide
just those and make that one a Build-dependency for the Slony-I package.

--
Oliver Elphick olly(at)lfix(dot)co(dot)uk
Isle of Wight http://www.lfix.co.uk/oliver
GPG: 1024D/A54310EA 92C8 39E7 280E 3631 3F0E 1EC0 5664 7A2F A543 10EA
========================================
"I press toward the mark for the prize of the high
calling of God in Christ Jesus."
Philippians 3:14

On Sat, Oct 23, 2004 at 04:59:26PM +0100, Oliver Elphick wrote:
>
> I thought someone was working on making a subsection of postgresql that
> would be sufficient to let packages like Slony-I build independently; is
> that making any progress?

Chris recently said on the slony lists that he's aiming to do this
for 1.1, and I have reason to believe that that'll be well on the
road to release by the end of the year. Chris, are those bits
ready? Maybe Oliver could use them as inspiration for the Debian
work?

A

--
Andrew Sullivan | ajs(at)crankycanuck(dot)ca
The fact that technology doesn't work is no bar to success in the marketplace.
--Philip Greenspun