taimaz/cdo repository overview

⁠Optimized CDO 2.5.4 (Climate Data Operator)

This image provides a high-performance build of Climate Data Operators (CDO) specifically tuned for massive datasets.

⁠Why This Image?

Standard CDO installations often use a non-thread-safe HDF5 backend. This leads to Segmentation Faults or NetCDF: HDF error when performing complex piped operations (e.g., mergetime | splityearmon) on large files (100GB+).

This build solves those issues by enforcing thread-safety at the library level.

⁠Key Components

• CDO: 2.5.4 (Latest Stable)
• HDF5: 1.14.6 (Compiled with --enable-threadsafe)
• NetCDF-C: 4.9.3
• ecCodes: 2.41.0 (GRIB1/2 support)
• Libraries: Includes SZIP (libaec), PROJ, and UDUNITS2.

⁠Usage Example

For processing large-scale data (e.g., 0.5 TB), use the following recommended flags:

docker run --rm \
    --shm-size=2gb \
    --ulimit nofile=10240:10240 \
    -v $(pwd):/data \
    taimaz/cdo:latest \
    -L -P 8 -f nc4 -z zip_4 \
    splityearmon -mergetime "*.nc" "output_"