Currently OSGeo has two machines hosted at Peer1. One is the machine holding
the drupal instance, svn, trac, ldap, etc and the other, as I understand it,
is essentially a frequently updated backup of the main machine that could
be used as a backup. Is that right?
My understanding is that some of us have been using this secondary machine
as a place to test changes to drupal, ldap, etc. Is this appropriate? Isn't
there a risk that the backup will be botched when we are actually need it?
If we shouldn't be using it as a development server, should I look into getting
the dev instance described below updated to match what is on www.osgeo.org so
it can be used for testing?
Kanhaiya - can you comment on what challenges there would be in duplicating
the main server as closely as possible on this dev server?
Best regards,
--
---------------------------------------+--------------------------------------
I set the clouds in motion - turn up | Frank Warmerdam, warmerdam@pobox.com
light and sound - activate the windows | http://pobox.com/~warmerdam
and watch the world go round - Rush | President OSGeo, http://osgeo.org
Yes the second server was/is meant to be the backup server if anything
happens to the main server. Right now it holds the backups of the main
server and there is still some work for this machine to automagically
'mirror' the main.
The reason we set up a test instance of drupal on the second machine
over the holidays - we had accesss to it and it's identical to the
production server for testing.
shawn
Frank Warmerdam wrote:
Shawn, etc,
Currently OSGeo has two machines hosted at Peer1. One is the machine
holding
the drupal instance, svn, trac, ldap, etc and the other, as I understand
it,
is essentially a frequently updated backup of the main machine that could
be used as a backup. Is that right?
My understanding is that some of us have been using this secondary machine
as a place to test changes to drupal, ldap, etc. Is this appropriate?
Isn't
there a risk that the backup will be botched when we are actually need it?
If we shouldn't be using it as a development server, should I look into
getting
the dev instance described below updated to match what is on
www.osgeo.org so
it can be used for testing?
The reason we set up a test instance of drupal on the second machine
over the holidays - we had accesss to it and it's identical to the
production server for testing.
Does this include mirroring latest additions that we've made like installing a newer version of Trac and Subversion? Or is it that the two servers started at the same point and then diverged as we've added to the production box?
I did install the new subversion on the second server but, i haven't
installed trac. Other than that I believe the 2 servers are identical.
shawn
Howard Butler wrote:
At 02:51 PM 1/31/2007, shawn barnes wrote:
The reason we set up a test instance of drupal on the second machine
over the holidays - we had accesss to it and it's identical to the
production server for testing.
Does this include mirroring latest additions that we've made like
installing a newer version of Trac and Subversion? Or is it that the
two servers started at the same point and then diverged as we've added
to the production box?
Yes the second server was/is meant to be the backup server if anything
happens to the main server. Right now it holds the backups of the main
server and there is still some work for this machine to automagically
'mirror' the main.
The reason we set up a test instance of drupal on the second machine
over the holidays - we had accesss to it and it's identical to the
production server for testing.
Shawn,
Does that means we don't really have any sort of period mirroring to the
second server happening yet?
By the way, the notes on the peer1 server setup are looking good!
Best regards,
--
---------------------------------------+--------------------------------------
I set the clouds in motion - turn up | Frank Warmerdam, warmerdam@pobox.com
light and sound - activate the windows | http://pobox.com/~warmerdam
and watch the world go round - Rush | President OSGeo, http://osgeo.org
I have quite a bit more documentation that i will be adding over the
next little while...probably a long while as we work things out on the
servers. And definitly adding a section on dns.
What is happening right now on the mirroring is just an rsync of all
backups from the main server to the second server. I have manually
mirrored - to figure out a good process not as a task - not planning on
being a human cron job :). and i knew there was some drupal testing so
i didn't want to get in the way of drupal ldap testing.
I do plan on the second server being a complete mirror. backing up to it
more than once is day is preferred and then using the backups to
automate the update.
shawn
Frank Warmerdam wrote:
shawn barnes wrote:
Frank,
Yes the second server was/is meant to be the backup server if anything
happens to the main server. Right now it holds the backups of the main
server and there is still some work for this machine to automagically
'mirror' the main.
The reason we set up a test instance of drupal on the second machine
over the holidays - we had accesss to it and it's identical to the
production server for testing.
Shawn,
Does that means we don't really have any sort of period mirroring to the
second server happening yet?
By the way, the notes on the peer1 server setup are looking good!
I did install the new subversion on the second server but, i haven't
installed trac. Other than that I believe the 2 servers are identical.
Shawn,
But things are just manually updated? For instance, if I edit /etc/aliases
on the main server, there is nothing propogating those changes to the backup,
right?
It seems like if we are serious about mirroring we need a more comprehensive
approach to it.
Best regards,
--
---------------------------------------+--------------------------------------
I set the clouds in motion - turn up | Frank Warmerdam, warmerdam@pobox.com
light and sound - activate the windows | http://pobox.com/~warmerdam
and watch the world go round - Rush | President OSGeo, http://osgeo.org
Hi Frank,
I got the problem. yes, we can do automated process for file synchronisation.
Actually “rsync” will take care of file synchronisation of two servers (one is main server and another is backup/testing server).
But i have to see for mysql database synchronisation. Currently im not aware of mysql database synchronisation.
I will see and let you know.
If you think that we should proceed for file synchronisation atleast, then let us know, so that we can start work according to that.
Currently OSGeo has two machines hosted at Peer1. One is the machine holding
the drupal instance, svn, trac, ldap, etc and the other, as I understand it,
is essentially a frequently updated backup of the main machine that could
be used as a backup. Is that right?
My understanding is that some of us have been using this secondary machine
as a place to test changes to drupal, ldap, etc. Is this appropriate? Isn’t
there a risk that the backup will be botched when we are actually need it?
If we shouldn’t be using it as a development server, should I look into getting
the dev instance described below updated to match what is on www.osgeo.org so
it can be used for testing?
Kanhaiya - can you comment on what challenges there would be in duplicating
the main server as closely as possible on this dev server?
Best regards,
---------------------------------------±-------------------------------------
I set the clouds in motion - turn up | Frank Warmerdam, warmerdam@pobox.com
light and sound - activate the windows | http://pobox.com/~warmerdam
and watch the world go round - Rush | President OSGeo, http://osgeo.org