[ZODB-Dev] zeo 2 availability, plans?

Mark McEahern mark@mceahern.com
Fri, 30 Aug 2002 11:42:29 -0500


[Jeremy Hylton [mailto:jeremy@zope.com]]
> I think you're concluding too much from what some old web pages say
> <0.5 wink>.  The ZEO FAQ is more than 2 years old.  (I didn't even
> know it existed.)  When we started the ZEO2 project, we rather loosely
> described the project as having two phases.  Actual development never
> really followed the plan.

Yeah, I figured that FAQ was probably out of date, but there wasn't anything
more recent as far as I could tell so I just hoped and prayed my illusions
would be shattered.  <wink>

I have two followup questions (probably perfect candidates for a FAQ):

Q:  What is the migration path from a standalone normal Zope installation to
a ZEO installation.  Would I be naive and foolish to think I can just take
the Data.fs, move it to the ZEO server, modify the Zope startup to talk to
ZEO, and all would be well?  Is the typical route to do something like
export the root from standalone Zope, import the root (over the top of? /
into?) into ZEO?  I'm happy to go about this in a trial and error sort of
way (which is inevitable anyway), but it'd sure be great to have some
guidance, suggestions on this.

Q:  Does ZEO2 address the lack of authentication in ZEO1?

> 2.0 final is around the corner.  I expect we'll do one more beta and
> get the final release out sometime in September.  This isn't a
> timetable, just my guess about how much time we'll have to work on it.

Awesome news!

[snip]
> Zope Corp's Zope Replication Service (ZRS) product address the
> reliability of the storage server.  A very high-level summary is: ZRS
> extends the ZEO architecture to ensure that every transaction written
> to a primary storage is replicated to any number of secondary
> storages.  These secondary storages can reside on different machines,
> and even different networks, to guard against hardware and network
> failures.  The data in a secondary storage is a logical copy of the
> primary storage.  Updates are replicated asynchronously to maximize
> performance and minimize data loss if the primary fails.
>
> Feel free to send me private email if you want to know more about ZRS.

Thank you Jeremy, we'll probably try a low-tech approach, but I'm glad to
know there's a product available for this as well.

Cheers,

// mark

-