Towards Near-Data Processing in Deep and Cold Storage

深低温存储中的近距离数据处理

展开查看详情

1.DLR.de • Chart 1 > XLDB’19 > Lightning Talk > Marcus Paradies > 03.04.2019 Towards Near-Data Processing in Deep and Cold Storage Hierarchies Marcus Paradies, German Aerospace Center (DLR) 04/03/2019, XLDB

2.DLR.de • Chart 2 > XLDB’19 > Lightning Talk > Marcus Paradies > 03.04.2019 The Storage Hierarchy and Its (Current) Coverage in DB Research DNA Tape Disk Flash Memory LLC Offline/ Nearline/ Online Nearline Offline

3.DLR.de • Chart 3 > Lecture > Author • Document > Date

4. DLR.de • Chart 4 > Lecture > Author • Document > Date Active Data Archives for Scientific Data

5.DLR.de • Chart 5 > XLDB’19 > Lightning Talk > Marcus Paradies > 03.04.2019 Scientific Application Domains Earth Observation Radio Astronomy Weather Forecasting Archive: 14 PB Archive: 50 PB Archive: 100 PB Disk Cache: 175 TB Disk Cache: 750 TB Disk Cache: 1.34 PB

6.DLR.de • Chart 6 > XLDB’19 > Lightning Talk > Marcus Paradies > 03.04.2019 Data Movement as Major Performance Bottleneck Active data archives (and their catalogs) are like Amazon, but just for data. No SLAs on access latency, usually between minutes and hours Tail latency can be multiple days Historic data analysis can easily request 100s of TB … Compute Disk Cache N Disk Cache 1 Data Archive Facilities

7.DLR.de • Chart 7 > XLDB’19 > Lightning Talk > Marcus Paradies > 03.04.2019 CryoDrill---Near-Data Processing for Cold Storage Focus on nearline storage (archival disks, tape) I‘d like to encourage everyone to not forget the nearline storage Consider layer! It‘s all not NDP opportunities considered (in- research topic, but it has a high a sexy network, in-storage) potential impact and is fun to work on! Push data reduction ops down the storage hierarchy