scalability   37781

« earlier    

Scaling Erlang cluster to 10,000 nodes
Maxim Fedorov - Scaling Erlang cluster to 10,000 nodes | Code Mesh LDN 18
erlang  scalability  scale  whatsapp 
7 days ago by clehene
The Mobile Network » Rakuten makes a real racket about its virtual network at MWC
How many radio heads will you install?
“If you look at the scale, roughly speaking 37,000 eNodeB times three [111,000] is the amount of radio heads we will have just for LTE only. It’s very large.”
rakuten  oran  vran  5g  ovum  scalability 
11 days ago by yorksranter
How Pusher Channels has delivered 10,000,000,000,000 messages - Making Pusher
On Friday night, at 22:20 UTC, Pusher Channels reached 10,000,000,000,000 total messages delivered. Here's how we did it.
redis  sharding  scalability  fanout 
13 days ago by jonasbehmer
Applied Monotonicity: A Brief History of CRDTs in Riak
In terms of monotonic programming and leveraging CRDTs as part of a programming model, there’s plenty of work as well. The CALM conjecture (J. M. Hellerstein and Alvaro 2019) first made the connection between consistency and monotonic programming. Bloom (Alvaro et al. 2011) is a distributed programming model inspired by Datalog that provides monotonic programming over sets; BloomL (Conway et al. 2012) extends this model for lattice-based programming. Lasp (Meiklejohn and Van Roy 2015) is a functional programming model over CRDTs from the SyncFree group. DataFun (Arntzenius and Krishnaswami 2016) is a functional Datalog where all operations are monotonic. Finally, the high-performance Anna KVS (Wu et al. 2019) uses monotonicity internally.

At Carnegie Mellon University, in the Composable Systems lab, we have been working on a type system for verifying monotonicity of functions, that can be used to construct CRDTs from the ground up by composing together monotone (or antitone) functions. Our work is inspired by a problem we observed in LVars (Kuper and Newton 2013), Lasp, and BloomL: functions must be monotonic (or homomorphic, a special case of monotonicity) to ensure both correctness and convergence of the system; in each of these systems functions are assumed to be correctly implemented and annotated accordingly. As we demonstrated in this article, this is very difficult to get right. Not only is monotonicity difficult to get right in practice, for systems that use monotonicity, they must be high-performance and cheap. In the case of CRDTs, the Observed-Remove Set was much too expensive to actually use in practice, but it’s easy to implement and reason about. For any monotonic solution to gain adoption, it essential that the solutions be both easy to use and not prohibitively expensive.
riak  concurrency  scalability 
17 days ago by janpeuker
Scaling Event Sourcing for Netflix Downloads
Phillipa Avery and Robert Reta describe how Netflix successfully launched their Download feature with the use of a Cassandra-backed event sourcing architecture. They describe their event store implementation and cover what they learned along the way, and what they could have done better. Finally, they review some improvements and extensions that they are planning to address going forward.
es  eventsourcing  cqrs  scalability  architecture 
17 days ago by dewe
awesome-scalability
The Patterns Behind Scalable, Reliable, and Performant Large-Scale Systems :wave: https://twitter.com/top_sde - binhnguyennus/
architecture  performance  development  scalability 
20 days ago by tedw

« earlier    

related tags

!starred  201901  5g  ads  advocacy  alogirthm  amazon  architecture  article  aurora  automation  autoscale  autoscaling  availability  aws  aws_cost_management  azurefunctions  benchmarking  bigdata  blockchain  blog  bubble2  business  cache  cap  ch1  ch3  cli  cloud  cloudcomputing  cluster  code  concurrency  container  cookies  cqrs  crap  cryptocurrency  culture  data  database  databases  db  dba  design  developer  development  devops  dgraph  distcomp  distributed-consensus  distributed-systems  distributed_systems  distributedcomputing  distributedsystem  docker  dynamodb  ebook  ecosystem  ecosystems  ecs  elasticsearch  elixir  elk  embedded  erlang  es  esb6  eventing  events  eventsourcing  expensivecloud  facebook  fanout  favorites  freebase  github  go  golang  google  graphd  graphs  growth  guidelines  ha  hadoop  hierarchy  high-availability  high  howto  indexing  infrastructure  jameshamilton  js  json  kernel  klog  knowledge-base  knowledge  knowledgegraph  load-testing  loadbalancing  log-processing  logging  management  messagequeue  meta  metaweb  meteor  microsoft  monitoring  mysql  netflix  networking  nodejs  notes  npm  ondemand  online  operating  ops  optimization  oran  organizing  ovum  papers  parsing  patterns  pdf  perf  performance  pinboard  postgres  postgresql  privacy  productivity  programming  query  queue  rabbitmq  rakuten  rant  redis  reference  replication  report  research  riak  rust  scale  scaling  scheduling  search  security  server  serverless  setup  sharding  shell  signalr  sql  stack  startup  state  streaming  switzerland  systemdesign  tech  techarch  technology  tips  tool  toread  triples  tuning  upenn  vcs  video  vignette  vignettes  vran  wal  web  web_design  webdev  websocket  websockets  whatsapp 

Copy this bookmark:



description:


tags: