Glossary of Terms¶
- chunk
Each dataset is consisting of several chunks that store objects whose hash id starts with the same characters (and matches the chunk’s chunk id).
- chunk count
the total number of chunks in a dataset, valid numbers: 16, 256, 4096.
- chunk id
1-2-3 character hexadecimal number in string format depending on the chunk count of a dataset. All objects whose hash is starting with the chunk id must be in the chunk identified by the chunk id.
- delayed
At execution, each task is pickled, and a copy of it is created in the thread. If a class cannot be pickled, it cannot be a member of the task before execution.
Delayed
solves this problem by delaying the creation recursively until the execution in the thread starts- hash
40 digit hexadecimal value generated with SHA1 method.
- hash id
fields of a domain object can be used to find out which chunk the object will be in a dataset.
- id
fields that uniquely identify an object. Two objects deemed equal if their id’s match.