Known limitations

Time zone

We do not recommend changing the time zone on your server. The server is currently set to use local time and changing time zones will result in a mismatch between the original data import times stored in the server and the way the clients report them.

Too many open file descriptors

Starting with OMERO 5, the server works directly from original files. At times, this requires a significant number of open file handles. If you are having problems with large or frequent imports, are seeing “Too many open file descriptors” or similar, you may need to increase the maximum number of open files per process. On Linux, this may be done by setting the nofile limit in /etc/security/limits.conf, for example:

omero  soft  nofile  10000
omero  hard  nofile  12000

This permits the omero user to have 10000 open files per process, which may be increased up to a maximum of 12000 by the user. The username and limits will need adjusting for the specifics of your installation and usage requirements. Note that these settings take effect only for new logins, so the server and the shell or environment the server is started from will require restarting. Run ulimit -a as the user running OMERO to verify that the changes have taken effect.

Changing group permissions

If a group contains a projection made by one member from data owned by another user, you cannot make the group into a private group.

File format support

Large images

When you import an image over a certain size, OMERO will generate a pyramid of lower resolution images if it doesn’t already exist in the file. The threshold size is configurable using omero.pixeldata.max_plane_height and omero.pixeldata.max_plane_width but set to 3192x3192 pixels by default. However, this process can be very resource-intensive, depending on the size of the image as well as the image format and any data compression used, for example see PixelData threads and pyramid generation issues.

Further, OMERO never generates pyramids for large floating-point pixel type images.

For large floating-point images, follow the recommended workflow to have the best experience (see below).

The OMERO pyramid generation process should be considered as deprecated and instead it is recommended that users avoid these issues by converting their data to pyramidal OME-TIFF files before importing into OMERO. A number of suitable tools are available such as NGFF-Converter, bioformats2raw & raw2ometiff, bfconvert, Kheops, tifffile, aicsimageio, libvips and QuPath.

Large floating-point images

Pyramids of image tiles are currently not generated for large floating-point pixel type images.

This primarily affects the following file formats:

• Gatan DM3

• MRC

• TIFF

However, in some cases, the floating-point images without pyramids can be viewed in OMERO clients at full resolution (if the images are not too large).

This behaviour is configured by omero.pixeldata.max_plane_float_override. By default (True), OMERO overrides the requirement for floating-point images above the omero.pixeldata.max_plane_height and omero.pixeldata.max_plane_width to have pyramids, which allows them to be treated as regular images and possibly viewed in the clients.

However, this also allows OMERO to attempt the calculation of minimal and maximal pixel intensity for these images (normally disabled for large images because it is resource intensive to read every pixel value).

When the omero.pixeldata.max_plane_float_override is set to False on your server, OMERO will not attempt to treat large floating-point images as if they are smaller images, so any large images without pre-generated pyramids will not be viewable. However, this will protect the server from expensive attempts to calculate min/max pixel values.

It is recommended to pre-generate pyramidal OME-TIFF images as described above and to set omero.pixeldata.max_plane_float_override to False on your server.

Import of OME-NGFF

The import of OME-NGFF is currently limited to the command-line (CLI) importer only.

Naming of OME-NGFF images in OMERO

The default naming of the OME-NGFF Images imported into OMERO is not intuitive at the moment. Use the omero import -n option to achieve explicit naming.

Depth of scanning prior to import

The import might fail if the hierarchy of folders is exceeding the depth of scanning (default: 4). For formats using deeper hierarchy of folders such as OME-NGFF use omero import --depth option to set the depth of scanning of 10 (or more if necessary).

Calculation of minima and maxima pixel values

If images are imported with one of the omero import --skip options skipping calculation of the global minima and maxima pixel values, OMERO clients will use the extrema of the pixel type range by default. Users can adjust the minima/maxima via the rendering settings. Recalculating minima and maxima pixel values after import is currently not supported.

Flex data in OMERO.tables

If you are using the advanced configuration setting FlexReaderServerMaps for importing Flex data split between multiple directories for use with OMERO.tables, you should not upgrade beyond 5.0.x. Neither the 5.1 line nor OMERO 5.2 support this functionality.

LDAP

Enabling synchronization of LDAP on user login may override admin actions carried out in the clients, see Synchronizing LDAP on user login for details.