Downloaded 44 times
Uploaded byGilWon Oh
Cache in API Gateway
This document provides an overview of caching strategies when using an API gateway. It first discusses different types of caches like HTTP caches, DNS caches, and database caches. It then focuses on caching strategies for web services, specifically caching within an API gateway. It discusses patterns like cache-aside, read-through, and write-through. It also covers using replication caches for API gateway infrastructure data and distributing caches between nodes. Finally, it provides examples of caching technologies that could be used in an API gateway like Ehcache, Infinispan, Hazelcast, and Redis.
In this document
Powered by AI
Introduction of Gilwon Oh and date (2015.04.17) of the presentation.
Details about the presenter's responsibilities as a full stack developer with contact info.
Engaging in a fun guess game leading to topics on Cache and API Gateway.
Rhetorical questions about time to walk from Seoul to Busan/New York and data download estimations.
Basic distance and speed measurements including units like AU and light speed, along with cable maps.
Speed conversion data illustrating different time units up to milliseconds and the concept of time in broader terms.
Introduction to various caching types including HTTP, CPU, and Database Cache, focusing on WebService Cache.Different cache handling patterns like Cache Aside, Write-behind, and Read Through with a focus on API Gateway.
Introduction to JGroups and its architecture for building reliable group communication applications.
Thanking the audience for their attention.
More Related Content
![[야생의 땅: 듀랑고] 서버 아키텍처 Vol. 2 (자막)](https://cdn.slidesharecdn.com/ss_thumbnails/vol2-160427160825-thumbnail.jpg?width=600ounds&width=560&fit=bounds)
![[115]쿠팡 서비스 클라우드 마이그레이션 통해 배운것들](https://cdn.slidesharecdn.com/ss_thumbnails/115coupang-181011031522-thumbnail.jpg?width=600ounds&width=560&fit=bounds)
![[MLOps KR 행사] MLOps 춘추 전국 시대 정리(210605)](https://cdn.slidesharecdn.com/ss_thumbnails/mlops-basic-210605064957-thumbnail.jpg?width=600ounds&width=560&fit=bounds)
![Payment Gateway 생존전략 [KCP]](https://cdn.slidesharecdn.com/ss_thumbnails/kcp120206-120705035731-phpapp02-thumbnail.jpg?width=600ounds&width=560&fit=bounds)
![[Spring Camp 2013] Java Configuration 없인 못살아!](https://cdn.slidesharecdn.com/ss_thumbnails/javaconfiguration-140128013032-phpapp02-thumbnail.jpg?width=600ounds&width=560&fit=bounds)
![[인모비 글로벌 리포트] 인모비 인도네시아 사용자들의 일용 소비재와 생필품 모바...](https://cdn.slidesharecdn.com/ss_thumbnails/random-150624233939-lva1-app6891-thumbnail.jpg?width=600ounds&width=560&fit=bounds)
![[한국핀테크포럼] 회원사소개: 인터페이](https://cdn.slidesharecdn.com/ss_thumbnails/150515-150904093340-lva1-app6891-thumbnail.jpg?width=600ounds&width=560&fit=bounds)
![[Demo session] 관리형 Kafka 서비스 - Oracle Event Hub Service](https://cdn.slidesharecdn.com/ss_thumbnails/oraclouddevmeetupeventhubpublic-180521084803-thumbnail.jpg?width=600ounds&width=560&fit=bounds)
![[Hands-on] CQRS(Command Query Responsibility Segregation) 와 Event Sourcing 패턴 실습](https://cdn.slidesharecdn.com/ss_thumbnails/cqrsesv201804221730publish-180422082848-180423040836-thumbnail.jpg?width=600ounds&width=560&fit=bounds)
Cache in API Gateway
- 1.
- 2. 책임 FullStack 개발자 지향 https://medium.com/@kobi97 kobi97@gmail.com
- 3. Guess? (ForFun) Cache API Gateway Cache In API Gateway Conclusion & Useful Links
- 4. Guess? (ForFun) Cache API Gateway Cache In API Gateway Conclusion & Useful Links
- 5. 서울~부산(뉴욕) 까지 걸어서 몇시간 걸릴까요?
- 6. 서울 에서 뉴욕PC의 사진을 1,000장 다운로드 하는데 얼마나 걸릴까요?
- 7. AWS EC2 JPTokyo에 설치를 하면 미국 West, East 어느 Zone에 설치하는 것이 더 빠를까요?
- 8. AWS EC2의 JP<-> JP, JP <-> Singapore, JP <-> US 각각의 경우에 Delay는 얼마 정도 생길까요?
- 9. 우리 나라에서 해외로연결된 해저 광 케이블 망이 있는 곳 은?
- 10. 나의 Local Disk에서데이터를 읽는 것과 vs 내 PC에서 일본 PC Memory 안에 있는 데이터를 읽으면 대략 속도 차이가 얼마 정도 날까요?
- 11. 도로원표: 도시간의 거리를측정할 때는 사용
- 12. 천문단위(AU): 지구에서 태양까지의거리 (약 150,000,000 Km) 지구 한 바퀴: 약 40,000 km (반지름: 6,400km) 지구에서 달: 약 384,400 km 광속 : 진공 상태에서 빛이 나아가는 속도 (약 300,000 km/s) 음속: 시속 1,224km 음속이 초당 340미터(섭씨 15도) 마하: 소리의 속도에 비해서 얼마나 빠른지 http://study.zum.com/book/13481
- 13.
- 14.
- 15. https://twitter.com/rzezeski/status/398306728263315456 (*1 ms: 10−31 μs: 10−6 1 ns: 10⁻⁹ Millennia(Millennium): 1000 years )
- 16. Guess? (ForFun) Cache API Gateway Cache In API Gateway Conclusion & Useful Links
- 18. DBWASClient WEB HTTP Cache .Browser caches . Proxy caches . Gateway caches(reverse proxy cache) DNS Cache … CPU Cache Page Cache Disk Cache ... DB Internal Cache Focus on WebService Cache Especially API Gateway *Java Cache 표준 (Jboss inifinispan 주도) JSR 107 : JCACHE - Java Temporary Caching API JSR 347 : Data Grids for the Java platform (2015.1 withdrawal OTL)
- 19. Replication Cache Invalidation Cache Distribution +L1 Cache Distribution Cache Local Cache http://www.slideshare.net/opennaru/in-memory-data-grid-infinispanjboss-data-grid Focus on Distribution+ Local Cache 유지 Local Cache+ 데이터 변경시 다른 서버 동일키 삭제 Cache 공유 X 모든 Cache 동일 복제 Cache 1/n으로 분산, Cache 없으면 다른 Node에 물어봄 General Hash (ex: Elasticsearch shards) vs Consistent Hashing (ex: infinspan, memcache)
- 20. Cache Aside Pattern Client가R/W Storage 책임 Storage Client Cache Write-behind Pattern 비동기적으로 Storage Write (보통 delay 세팅 시간 후) Storage Client Cache Read Through/ Write Through Cache가 R/W Storage 책임 Storage Client Cache Refresh-Ahead Pattern 주기적으로 최근에 접근 한 cache Data를 Reload Storage Client Cache http://www.slideshare.net/tmatyashovsky/from-cache-to-in-memory-data-grid-introduction-to-hazelcast http://docs.oracle.com/cd/E15357_01/coh.360/e15723/cache_rtwtwbra.htm#COHDG5181
- 21. Guess? (ForFun) Cache API Gateway Cache In API Gateway Conclusion & Useful Links
- 22. API-G #1 API-G #2 4.Check Throttling & API Burst 1. API Call 2. API Validation 3. 2 Auth. 5. Mediation Service #1 Service #2 Service #3 …… 6. Target Service Call Storage #1 Storage #2
- 23. Interface WAS #1 API-G WAS #1 API-G WAS#2 Interface WAS #2 Use Remote Cache 4. Check Throttling & API Burst 1. API Call 2. API Validation 3. 2 Auth. 5. Mediation Service #1 Service #2 Service #3 …… 6. Target Service Call Management Center #1,2 Internal Interface API Use Flow CRUD: API Infra Data How To use cache? DB DB Storage #1 Storage #2
- 24. Guess? (ForFun) Cache API Gateway Cache In API Gateway Conclusion & Useful Links
- 25. Interface WAS #1 Cache Interface WAS #2 Cache CacheReplication Use Remote Cache Management Center #1,2 Internal Interface API Use Flow CRUD: API Infra Data DB DB API-G WAS #1 Cache API-G WAS #2 Cache Redis (Slave) Redis (Master) 4. Check Throttling & API Burst 1. API Call 2. API Validation 3. 2 Auth. 5. Mediation Service #1 Service #2 Service #3 …… 6. Target Service Call Use Ehcache-RMI P2P Use Redis-Sentinel
- 26. JPN USA Use ReplicationCache? Infinispan vs Hazlecast vs Ehcache-jGroups
- 27. Gossip Router #1 Gossip Router #2 JPNUSA Use Ehcache-JGroups Ehcache-JGroups Maybe almost same as Infinispan library mode
- 28. Cache in APIGateway |- Infra Data(Replication Cache) |- Ehcache-RMI |- Ehcache-JGroups (*Infinspan) |- API Burst & Throttling |- Redis, Redis-Sentinel(*Redis-Cluster) |- Http Cache (SKIP)
- 29. https://tedwon.atlassian.net/wiki/display/SE/Ehcache+Replicated+Cache http://www.slideshare.net/HyeonSeokChoi/overview-of-the-ehcache http://www.nextree.co.kr/p3151/ (아키텍처 설계:분산캐시-EhCache+TerracottaServer Array) http://www.slideshare.net/alexmiller/scaling-your-cache http://ehcache.org/documentation/2.8/get-started/getting-started
- 30. http://www.slideshare.net/charsyam2/webservice-cache-strategy http://www.slideshare.net/charsyam2/cache-governancepub http://www.slideshare.net/charsyam2/redis-edu-1 http://www.slideshare.net/charsyam2/redis-edu-2 http://www.slideshare.net/charsyam2/redis-edu-3 http://www.slideshare.net/charsyam2/redis-edu-4 http://www.slideshare.net/charsyam2/redis-edu-5 http://www.slideshare.net/krisjeong/this-is-redis-kor-slideshare http://www.slideshare.net/krisjeong/redis-data-design-by http://www.joinc.co.kr/modules/moniwiki/wiki.php/man/12/REDIS/DataModeling http://www.slideshare.net/Byungwook/redis-data-modeling-examples http://www.mimul.com/pebble/default/2012/09/28/1348812530141.html
- 31. http://www.slideshare.net/opennaru/in-memory-data-grid-infinispanjboss-data-grid http://bcho.tistory.com/m/post/883 http://www.slideshare.net/jaehongc/jco-2014?related=1(인피니스팬데이터그리드 따라잡기 (@JCO 2014) ) 이제 필요한 것은 In Memory Data Grid http://helloworld.naver.com/helloworld/106824 in memory computing http://blog.lgcns.com/176
- 32. https://labs.consol.de/java-caches (Ehcachevs Hazelcast vs Infinispan) http://www.slideshare.net/tmatyashovsky/from-cache-to-in-memory-data-grid- introduction-to-hazelcast http://www.slideshare.net/tmatyashovsky/distributed-applications-using- hazelcast?from_m_app=android http://blog.ragozin.info/2009/04/caches-are-everywhere.html http://cutajar.weebly.com/not-a-blog/caching-techniques 확장성 있는 웹 아키텍처와 분산 시스템 http://helloworld.naver.com/helloworld/textyle/206816 확장 가능한 분산 인 메모리 캐시를 제공하는 ElastiCache http://pyrasis.com/book/TheArtOfAmazonWebServices/Chapter15 분산 메모리기반 플랫폼 Plandas # Cache Cloud http://readme.skplanet.com/?p=8884 http://knowledgecloud.tistory.com/entry/EC2-에서-클러스터링-시-유연한-설정-가져가기
- 33.
- 34.
- 35. GROUP (within the sameLAN, or across a WAN) Process Process Process message Process join leave or crash message
- 36. It consists of3 parts: (1) Channel used by application programmers to build reliable group communication applications (2) Building blocks layered on top of the channel and provide a higher abstraction level (3) Protocol stack protocols implements the properties specified for a given channel. Application Building blocks JChannel GMS MERGE FRAG UDP/TCP Network Figure 1.1 the architecture of JGroups
- 37. Application Building blocks JChannel GMS MERGE FRAG UDP/TCP Network Figure 1.1the architecture of JGroups send() Message delivery receive() callback Network listen Transport protocol Topmost protocol connect /disconn. start /stop close destory (release) Message delivery
- 38. Protocol List Transport UDP,TCP, Tunnnel Initial membership discovery PING, TCPPING, TCPGOSSIP, MPING, FILE_PING, JDBC_PING, BPING, RACKSPACE_PING, S3_PING, SWIFT_PING, AWS_PING, PDC(Persistent Discovery Cache) Merging after a network partition MERGE2, MERGE3 Failure Detection FD, FD_ALL, FD_SOCK, FD_PING, VERIFY_SUSPECT Reliable Message Transmission Pbcast.NAKACK, NAKACK2, UNICAST, UNICAST2, UNICAST3, RSVP Message stability STABLE Group Membership Pbcast.GMS Flow control FC, MFC, UFC Fragmentation FRAG, FRAG2 Ordering SEQUENCE, Total Order Anycast(TOA) State Transfer Pbcast.STATE_TRANSFER, StreamingStateTransfer, pbcast.STATE, STATE_SOCK, BARRIER Pbcast.FLUSH Misc Statistics, Security, COMPRESS, SCOPE, RELAY, RELAY2, STOMP, DAISYCHAIN, RATE_LIMITER, Locking protocols, CENTRAL_EXECUTOR, COUNTER, SUPERVISOR
- 39. GROUP View api-g1-9235(coordinator) api-g2-32349 api-g3-23412 Channel (api-g1-29235) Channel (api-g2-32349) Channel (api-g3-23412) FD_SOCK FD VERIFY_SUSPECT GMS GMS VERIFY_SUSPECT FD FD_SOCK FD_SOCK FD VERIFY_SUSPECT GMS FD HeartBeat FDHeartBeat FD HeartBeat FD Sock Connection FD Sock Connection FailFD Sock Connection SUSPECT ping 1 2 3 4 view changing
- 40.