|
Home > Archive > Slony1 PostgreSQL Replication > August 2005 > master master asynchronous
You are viewing an archived Text-only version of the thread.
To view this thread in it's original format and/or if you want to reply to
this thread please [click here]
| Author |
master master asynchronous
|
|
| David Durham 2005-08-11, 1:28 pm |
| From the docs it looks like this is not supported by slony. Is that
correct? Is there anything for postgres that does support an
asynchronous master to master replication scenario?
Dave
| |
| Andrew Sullivan 2005-08-11, 1:28 pm |
| On Thu, Aug 11, 2005 at 09:57:15AM -0500, David Durham wrote:
> From the docs it looks like this is not supported by slony. Is that
> correct? Is there anything for postgres that does support an
> asynchronous master to master replication scenario?
Because Slony-I replicates by table, rather than by database, you can
do this _only if_ you have some intelligence in your application.
AFAIK, there is no stable multimaster async system available for
Postgres.
A
--
Andrew Sullivan | ajs-oaT0K0jot5/q2IAV+ODieA@public.gmane.org
When my information changes, I alter my conclusions. What do you do sir?
--attr. John Maynard Keynes
| |
| David Durham 2005-08-11, 1:28 pm |
| Andrew Sullivan wrote:
> Because Slony-I replicates by table, rather than by database, you can
> do this _only if_ you have some intelligence in your application.
I don't understand this. What kind of intelligence, conflict resolution?
Another question, that I haven't been able to find an answer for is
whether or not Slony-I replication supports LargeObjects.
BTW, is there a recommended searchable archive of this list?
Thanks,
- Dave
| |
| Darcy Buskermolen 2005-08-11, 1:28 pm |
| On Thursday 11 August 2005 10:00, David Durham wrote:
> Andrew Sullivan wrote:
>
> I don't understand this. What kind of intelligence, conflict resolution?
>
> Another question, that I haven't been able to find an answer for is
> whether or not Slony-I replication supports LargeObjects.
No Slony-I does not replicate large objects (bytea is supported fully though).
This is covered in the overview
http://gborg.postgresql.org/project...?howto_overview
>
> BTW, is there a recommended searchable archive of this list?
>
> Thanks,
>
>
> - Dave
> ____________________
____________________
_______
> Slony1-general mailing list
> Slony1-general- AuKwsB3Fm+ugFIWk8tvy
RWD2FQJk+8+b@public.gmane.org
> http://gborg.postgresql.org/mailman.../slony1-general
--
Darcy Buskermolen
Wavefire Technologies Corp.
http://www.wavefire.com
ph: 250.717.0200
fx: 250.763.1759
| |
| Andrew Sullivan 2005-08-11, 1:28 pm |
| On Thu, Aug 11, 2005 at 12:00:49PM -0500, David Durham wrote:
> Andrew Sullivan wrote:
>
>
> I don't understand this. What kind of intelligence, conflict resolution?
Well, that; but really, what you can do is something like this.
Suppose you have two offices, say CAnada and United States. And you
want to write into a table, activity.
You set up two tables, then, activity_ca and activity_us. The Canada
office database replicates activity_ca to the US office, and the US
office database replicates activity_us to the Canada office. And
"activity" is actually a view which is a UNION of the two.
You could, I'll bet, do this with rules on the view to make it
updateable, too; you'd need different rules in each database.
Your problem will be if the Canada office wants to make updates to
the US office table; you'll get a Slony error then.
> Another question, that I haven't been able to find an answer for is
> whether or not Slony-I replication supports LargeObjects.
Nope. Use bytea, which will work.
A
--
Andrew Sullivan | ajs-oaT0K0jot5/q2IAV+ODieA@public.gmane.org
A certain description of men are for getting out of debt, yet are
against all taxes for raising money to pay it off.
--Alexander Hamilton
| |
| Christopher Browne 2005-08-11, 1:28 pm |
| David Durham wrote:
> From the docs it looks like this is not supported by slony. Is that
> correct?
Correct.
> Is there anything for postgres that does support an asynchronous
> master to master replication scenario?
Rumor has it that PeerDirect has something of the sort; the model is
apparently somewhat akin to the "PalmPilot Sync", where each host is
potentially a relatively equal "peer" (hence the name!).
Sybase has a replication system that apparently works in a somewhat
similar fashion. At some point, I could see someone taking a version of
Slony-I and "forking" it to build an async multimaster system.
One trouble case that comes up really quickly is that of the handling of
"balance" columns. That is, things like account balances, where updates
are intended to *adjust* the balance, not to "set" the balance.
Obvious policies for automated conflict resolution systems don't do
terribly well with account balances :-(.
| |
| David Fetter 2005-08-11, 1:28 pm |
| On Thu, Aug 11, 2005 at 12:10:43PM -0400, Andrew Sullivan wrote:
> On Thu, Aug 11, 2005 at 09:57:15AM -0500, David Durham wrote:
>
> Because Slony-I replicates by table, rather than by database, you
> can do this _only if_ you have some intelligence in your
> application. AFAIK, there is no stable multimaster async system
> available for Postgres.
You could have left the last two words off that sentence without
reducing its validity :)
Cheers,
D
--
David Fetter david- iTBeR5ZXhegdnm+yROfE
0A@public.gmane.org http://fetter.org/
phone: +1 510 893 6100 mobile: +1 415 235 3778
Remember to vote!
| |
| Andrew Sullivan 2005-08-11, 1:28 pm |
| On Thu, Aug 11, 2005 at 11:19:52AM -0700, David Fetter wrote:
> On Thu, Aug 11, 2005 at 12:10:43PM -0400, Andrew Sullivan wrote:
>
> You could have left the last two words off that sentence without
> reducing its validity :)
Well, my understanding is that Sybase's system works, for some value
thereof. And people apparently do use it for production-grade
systems. Now, whether _I_ would do so is another matter.
But people really do seem to need to hear that multimaster
replication of any kind is a _hard problem_. This isn't something
you're gonna throw together in a day.
A
--
Andrew Sullivan | ajs-oaT0K0jot5/q2IAV+ODieA@public.gmane.org
This work was visionary and imaginative, and goes to show that visionary
and imaginative work need not end up well.
--Dennis Ritchie
| |
| David Fetter 2005-08-11, 8:25 pm |
| On Thu, Aug 11, 2005 at 02:22:48PM -0400, Andrew Sullivan wrote:
> On Thu, Aug 11, 2005 at 11:19:52AM -0700, David Fetter wrote:
>
> Well, my understanding is that Sybase's system works, for some value
> thereof. And people apparently do use it for production-grade
> systems. Now, whether _I_ would do so is another matter.
>
> But people really do seem to need to hear that multimaster
> replication of any kind is a _hard problem_.
Yes!
> This isn't something you're gonna throw together in a day.
*I* won't throw such a thing together, as I lack the needed hubris,
but Others do with depressing frequency and even more depressing
results.
> This work was visionary and imaginative, and goes to show that
> visionary and imaginative work need not end up well.
> --Dennis Ritchie
Heh.
Cheers,
D
--
David Fetter david- iTBeR5ZXhegdnm+yROfE
0A@public.gmane.org http://fetter.org/
phone: +1 510 893 6100 mobile: +1 415 235 3778
Remember to vote!
|
|
|
|
|