Details
-
Bug
-
Resolution: Fixed
-
Critical
-
Cheshire-Cat
-
7.0.0-5294
-
Untriaged
-
Centos 64-bit
-
-
1
-
Unknown
-
Magma-July-5-2021
Description
Steps to Repro:
- Created a 4 node cluster (Each node having bucket RAM QUOTA ==100MB)
- Loaded 11 million items(doc size = 128, Pure creates)
- To avoid "ServerOutOfMemoryException" slow doc ops were used for doc loading (~ 700 ops/sec)
- Magma : Disk size was 34.8 GB
- Did above experiment on CouchStore(~ 11.5 Million items): Disk Size was 2.5 GB(approx)
Attachments
Issue Links
- mentioned in
-
Page Loading...
For Gerrit Dashboard: MB-46915 | ||||||
---|---|---|---|---|---|---|
# | Subject | Branch | Project | Status | CR | V |
156526,5 | MB-46915 lsm: Remove minimum 2MB preallocation for data sstables | master | magma | Status: MERGED | +2 | +1 |
156850,9 | MB-46915 lsm: Redefine logical size computation for sstable | master | magma | Status: MERGED | +2 | +1 |
156851,11 | MB-46915 lsm: Introduce WastedSpace() API for sstable | master | magma | Status: MERGED | +2 | +1 |
156852,16 | MB-46915 magma: Introduce internal fragmentation based compaction | master | magma | Status: MERGED | +2 | +1 |
156905,15 | MB-46915 tests: Add test for internal fragmentation triggered compaction | master | magma | Status: MERGED | +2 | +1 |
157094,6 | MB-46915 lsm: Exclude compacting wasted space in computing level size | master | magma | Status: MERGED | +2 | +1 |