[ome-users] FW: OMERO to OMERO communications?

[John Webber (NBI)]

Hi Josh, 

Thank you for your quick reply.

The data we are looking to share between two different Omero systems will be for viewing. The syncing will be one-way only though - images will not be shared back to the first Omero server.  
	
This syncing will purely be from on source / internal server through to a destination, externally visible server.  In this I am also alluding to the driving force for this set up:  we would like one Omero server that is externally facing in our DMZ, but we need a mechanism for scientists here to share images with external collaborators, who would like to view these images (and, if possible, workflows!)

The two options I am currently evaluating are:

	(1)	As you mention below, exporting the data as OME-TIFF, from one Omero system and importing it into another.
	(2)	A mechanism to back up the source database and then restore this data to the other database regularly.

If you have any other ideas, or pointers, I would be extremely grateful.  Do you think I could look to achieve these objectives using Omero Scripts in some way?

Thanks
John



-----Original Message-----
From: Josh Moore [mailto:josh at glencoesoftware.com] 
Sent: 08 October 2013 10:08
To: John Webber (NBI)
Cc: ome-users at lists.openmicroscopy.org.uk
Subject: Re: [ome-users] OMERO to OMERO communications?


On Oct 7, 2013, at 5:29 PM, John Webber (NBI) wrote:

> Hi all,

Hi John,

> I am currently looking to implement some form of OMERO to OMERO communication, for sharing data between different Omero instances / servers.  I am specifically looking for a mechanism by which selected Omero images (not all) within one Omero server, could be made available to the database in a second Omero server.

Solely for viewing? Also for annotating, etc? What are the data sizes you're looking to "mirror"/"federate"/"share"? What are the driving forces for multiple servers as opposed to multiple groups?


> Does anyone have any pointers as to the best method that this can be achieved?


Options requiring little or no new development:

 * The current best-practice would be to export an OME-TIFF of each image and import it into the second system.

 * With images newly imported in OMERO5, you would be able to export the original files rather than just an OME-TIFF.


Other options:

 * A further possibility would be to point two OMERO5 servers at the same physical location. This would likely fall under "In place imports" recently discussed on the mailing list[1]

 * More invasive but still possible would be a mechanism that would pull the bytes from some other location. For example, a "RemotePixelBuffer" similar to "BfPixelBuffer" [2]. We've begun thinking about a solution like this, but it would undoubtedly be a good deal of work.


> Thanks
> John

Cheers,
~Josh


[1] http://lists.openmicroscopy.org.uk/pipermail/ome-devel/2013-September/002523.html
[2] https://github.com/openmicroscopy/openmicroscopy/blob/develop/components/romio/src/ome/io/bioformats/BfPixelBuffer.java