BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:America/Chicago
X-LIC-LOCATION:America/Chicago
BEGIN:DAYLIGHT
TZOFFSETFROM:-0600
TZOFFSETTO:-0500
TZNAME:CDT
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0500
TZOFFSETTO:-0600
TZNAME:CST
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20230124T171521Z
LOCATION:C148
DTSTART;TZID=America/Chicago:20221114T161500
DTEND;TZID=America/Chicago:20221114T163500
UID:submissions.supercomputing.org_SC22_sess460_ws_pdsw108@linklings.com
SUMMARY:DENKV: Addressing Design Trade-Offs of Key-Value Stores for Scient
 ific Applications
DESCRIPTION:Workshop\n\nDENKV: Addressing Design Trade-Offs of Key-Value S
 tores for Scientific Applications\n\nJamil, Khan, Kim, Lee, An...\n\nHPC f
 acilities have employed flash-based storage tier near to compute nodes to 
 absorb high I/O demand by HPC applications during periodic system-level ch
 eckpoints. To accelerate these checkpoints, proxy-based distributed key-va
 lue stores (PD-KVS) gained particular attention for their flexibility to s
 upport multiple backends and network configurations. PD-KVS rely internall
 y on monolithic KVS, such as RocksDB, to exploit the KV interface and quer
 y support. However, PD-KVS are unaware of high redundancy factor in checkp
 oint data, which can be up to GBs to TBs, and therefore, tend to generate 
 high write and space amplification on these storage layers. We propose DEN
 KV which is deduplication-extended node-local LSM-tree-based KVS. DENKV em
 ploys asynchronous partially inline dedup (APID) and aims to maintain the 
 performance characteristics of LSM-tree-based KVS while reducing the write
  and space amplification problems. We implemented DENKV atop BlobDB and sh
 owed that our solution maintains performance while reducing write and spac
 e amplification.\n\nSession Format: Recorded\n\nRegistration Category: Wor
 kshop Reg Pass
END:VEVENT
END:VCALENDAR