[LISPMAP-136] Expired mapping removed from map-cache, not removed from MD-SAL Created: 25/Aug/16 Updated: 19/Oct/17 Resolved: 06/Sep/17 |
|
| Status: | Resolved |
| Project: | lispflowmapping |
| Component/s: | Service |
| Affects Version/s: | Boron |
| Fix Version/s: | None |
| Type: | Bug | ||
| Reporter: | Lori Jakab | Assignee: | Unassigned |
| Resolution: | Done | Votes: | 0 |
| Labels: | None | ||
| Remaining Estimate: | Not Specified | ||
| Time Spent: | Not Specified | ||
| Original Estimate: | Not Specified | ||
| Environment: |
Operating System: All |
||
| External issue ID: | 6537 |
| Priority: | Normal |
| Description |
|
When a southbound originated mapping lookup triggers the lazy removal of an expired mapping, said mapping is only removed from the map-cache, but not from the operational MD-SAL data-store. |
| Comments |
| Comment by Sheikh Shakib Ahmed [ 30/Aug/16 ] |
|
I have gone through the mapRequest from Lisp Southbound services which ultimately calls getMapping(src, dst) and which calls getSbMappingWithExpiration(src, dst) which I think is the function being called to get a south bound mapping. If we look at the following piece of code, private Object getSbMappingWithExpiration(Eid src, Eid dst) { } This actually deletes the mapping from Data Store Back End but not from the map cache itself. |
| Comment by Vina Ermagan [ 01/Sep/16 ] |
|
Thi sis actually reversed: Mapping is removed from MD-SAL, but not from the local map-cache, and also not from other cluster nodes' map caches. |
| Comment by Vina Ermagan [ 06/Sep/16 ] |
|
Fix: https://git.opendaylight.org/gerrit/#/c/45121/ Pending review |
| Comment by Lori Jakab [ 06/Sep/16 ] |
|
No, the bug description is correct. The mapping is removed from the map-cache inside SimpleMapCache#getMappingLpmEid, on lines 177-178. However, 'MappingMergeUtil.mappingIsExpired(mapping)' no longer works, since we removed the timestamp from MappingRecord, so 'dsbe.removeMapping()' is not invoked. I have a different approach in mind, which will help with |
| Comment by Sheikh Shakib Ahmed [ 06/Sep/16 ] |
|
I have taken a deep look and Lori is right. I also have a solution in my thought and will love to discuss it with you. Please let me know when you are free. |
| Comment by Lori Jakab [ 12/Aug/17 ] |
|
Old bug, need to see if it is still valid. |
| Comment by Lori Jakab [ 06/Sep/17 ] |
|
Fixed by the time bucket patch series, verified manually. |