[ome-devel] Setting up an OMERO server for Aperio ScanScopeXT big images

[Kenneth Gillen]

Dear Luca,

We appreciate the issues you are having with Aperio/ScanScope images, and
unfortunately this will always be the case with the current version of
OMERO, the 4.4.x line.

Right now the OME team are in the last stages of releasing OMERO 5 [1],
which solves OMERO 4's current problems with "big images" such as these,
seeing a vast performance improvement here.

Over the next few months we will be producing Release Candidates of
OMERO-5, which are really the answer to your problem here. If possible,
hold out for these, though there are likely to be API and DB changes for
some time.

In the short term, there may be performance tweaks you can perform at the
moment on your laptop, such as allocating more memory to the JVM heap, as
documented at the openmicroscopy site [2]. This applies also if you still
plan to deploy 4.4.x to a server. The usual system requirements will apply
to your server, be it VM or physical [3] - a specific VM concern may be
how the storage is provisioned to the VM, which will depend on the
hypervisor you are deploying on. E.g. We have seen when using VirtualBox,
IO performance can be significantly faster when using a raw partition as
opposed to a VDI file-based disk image, on OS X.

If there is an extremely pressing need for working with big images and the
stable OMERO 4 codebase, there are some experimental configuration setups
which are possible, but are non-trivial, including using memecached for
caching which can help OMERO.web but not insight, though we don¹t have any
such configs documented or setups well-tried.

[1] https://trac.openmicroscopy.org.uk/ome/milestone/OMERO-5

[2]
https://www.openmicroscopy.org/site/support/omero4/sysadmins/server-advance
d-configuration.html
[3]
https://www.openmicroscopy.org/site/support/omero4/sysadmins/system-require
ments.html

Best,

Kenny


Kenneth Gillen

OME System Administrator

Wellcome Trust Centre for Gene Regulation & Expression
College of Life Sciences
MSI/WTB/JBC Complex
University of Dundee
Dow Street
Dundee  DD1 5EH
United Kingdom

Tel: +44 (0) 1382 386364
Skype: kennethgillen



------------------------------------
Luca Lianas luca.lianas at crs4.it
Tue Jun 11 16:55:54 BST 2013

Hi,
during the next week we are going to acquire a set of 400 images produced
using an Aperio ScanScopeXT (each image is about 1.5 GB of data and has a
resolution of about 100k x 50k pixels), I'm trying to import some test
images using my notebook as server just to check if everything works fine.
I'm able to import and see the images, the only "problem" I'm experiencing
is that the server is very slow during the image processing performed by
the Pixel-Service (I know that running both the server and the insight
client on a notebook with a 3 years old i5 processor is not the smartest
thing that I can do).
I'm planning to install a virtual machine as OMERO server, do you have any
advice about the VM configuration or the settings of the server to prevent
performance issues when uploading and viewing these kind of images?

Thanks,

Luca

--
Luca Lianas
Data Fusion - Distributed Computing
CRS4 - Centro di Ricerche, Sviluppo e Studi Superiori della Sardegna
POLARIS, Edificio 1, 09010 PULA (CA - Italy)


Phone: 070/9250266
Email: luca.lianas at crs4.it
<http://lists.openmicroscopy.org.uk/mailman/listinfo/ome-devel>




The University of Dundee is a registered Scottish Charity, No: SC015096