Details
-
Improvement
-
Status: Closed
-
Major
-
Resolution: Done
-
1.4.0
-
None
Description
The current architecture around the BLOB server and cache components seems rather patched up and has some issues regarding concurrency (FLINK-6380), cleanup, API inconsistencies / currently unused API (FLINK-6329, FLINK-6008). These make future integration with FLIP-6 or extensions like offloading oversized RPC messages (FLINK-6046) difficult. We therefore propose an improvement on the current architecture as described below which tackles these issues, provides some cleanup, and enables further BLOB server use cases.
Please refer to https://cwiki.apache.org/confluence/display/FLINK/FLIP-19%3A+Improved+BLOB+storage+architecture for a full overview on the proposed changes.
Attachments
Issue Links
- relates to
-
FLINK-5908 Blob Cache can (rarely) get corrupted on failed blob downloads
-
- Closed
-
-
FLINK-6380 BlobService concurrency issues between delete and put/get methods
-
- Closed
-
-
FLINK-6046 Add support for oversized messages during deployment
-
- Resolved
-
-
FLINK-7285 allow submission of big job graphs
-
- Closed
-
- supercedes
-
FLINK-1928 BlobManager can be a bottleneck on large deployments
-
- Closed
-
-
FLINK-2760 Avoid uploading of already existing jars from the JobClient
-
- Closed
-
-
FLINK-6329 refactor name-addressable accessors in BlobService classes from returning a URL to a File or InputStream
-
- Closed
-
- links to