Hot-Swapping Reference Data for Address Verification (On-Premises)

Hot-Swapping Reference Data for Address Verification (On-Premises)

During and After the Hot Swap

During and After the Hot Swap

Address Verification performs the following steps during a hot swap:
The following steps continue the example of swapping from
FileSetA
to
FileSetB
.
  1. Address Verification updates the contents of
    FileSetsInfo.json
    to reflect that the new file set is being prepared:
    { "FileSetAState": "InUse", "FileSetBState": "PreparingForUse" }
  2. Address Verification starts a new batch of function servers and preloads the reference data files in the
    FileSetB
    directory for the servers. During this time, the function servers that use
    FileSetA
    remain fully available for all jobs.
  3. When all new function servers have started successfully, the previous function servers are shut down.
  4. The contents of
    FileSetsInfo.json
    are updated to reflect that
    FileSetB
    is now in use and
    FileSetA
    has been unloaded:
    { "FileSetAState": "Unused", "FileSetBState": "InUse" }
As starting a new batch of function servers is akin to a regular initialization procedure, Address Verification writes messages logged during this sub-initialization to the initialization log. However, any error that prevents the hot swap from starting is written to the Address Verification error log.
After a successful hot swap, the reference data files that the operation unloaded are unlocked on the file system and can be moved or deleted again.

0 COMMENTS

We’d like to hear from you!