Project

General

Profile

Task #7708

Re-attempts of synchronization due to TransferObjectTask failures does not send SyncFailed to MN

Added by Robert Waltz about 8 years ago. Updated almost 8 years ago.

Status:
Closed
Priority:
Normal
Assignee:
Category:
d1_synchronization
Start date:
2016-04-05
Due date:
% Done:

100%

Story Points:
Sprint:

Description

The current code of V2TransferObjectTask will attempt to retry processing of a SyncObject task if a VersionMismatch exception is thrown.

After 5 attempts, the task will not be retried.

The number of retry attempts should be a configurable parameter and set as a global variable (get rid of it as a magic number).

The failure after 5 attempts should be reported back to the MN via a synchronization Failed message.

Associated revisions

Revision 17966
Added by Rob Nahf almost 8 years ago

refs: #7708. fixed VersionMismatch logic in processV1SystemMetadataUpdate

Revision 17966
Added by Rob Nahf almost 8 years ago

refs: #7708. fixed VersionMismatch logic in processV1SystemMetadataUpdate

History

#1 Updated by Rob Nahf almost 8 years ago

V2TransferObjectTask should only rely on VersionMismatch for objects under v1 systemMetadata control (Tier3 v1 nodes). The new commits (by me) that refactored retry handling messed with the VersionMismatch logic and categorized it as unrecoverable.

I restored the retry behavior for this exception, but also got rid of the special number of retries for this situation. Retry logic was getting too complicated, and the number of attempts was not split out by type of retry-able failure, so it seems pointless to try to differentiate the number of attempts allowed for a given "last-failure" situation.

#2 Updated by Rob Nahf almost 8 years ago

  • Status changed from New to Closed
  • % Done changed from 0 to 100
  • translation missing: en.field_remaining_hours set to 0.0

Also available in: Atom PDF

Add picture from clipboard (Maximum size: 14.8 MB)