The Morning Paper

FEBRUARY 6, 2020

Oh, and there’s a scheduler too of course to keep all the plates spinning. On the Cloudburst design teams’ wish list: A running function’s ‘hot’ data should be kept physically nearby for low-latency access. A low-latency autoscaling KVS can serve as both global storage and a DHT-like overlay network.

Serverless 98

Serverless Lambda Cache Latency 98

The Morning Paper

JANUARY 30, 2020

Edge servers are the middle ground – more compute power than a mobile device, but with latency of just a few ms. The client MWW combines these estimates with an estimate of the input/output transmission time (latency) to find the worker with the minimum overall execution latency.

Mobile 104

Mobile Cloud Latency Games 104

The Morning Paper

OCTOBER 26, 2020

It’s limited by the laws of physics in terms of end-to-end latency. We saw earlier that there is end-user pressure to replace batch systems with much lower latency online systems. We are observing significant demand from users in terms of avoiding batch telemetry pipelines altogether. Emphasis mine ). Emphasis mine ).

Cloud 104

Cloud Big Data Latency Architecture 104

All Things Distributed

OCTOBER 2, 2017

" Of course, no technology change happens in isolation, and at the same time NoSQL was evolving, so was cloud computing. Performant – DynamoDB consistently delivers single-digit millisecond latencies even as your traffic volume increases.

Internet 111

Internet Internet AWS Performance 111

All Things Distributed

DECEMBER 12, 2018

They can run applications in Sweden, serve end users across the Nordics with lower latency, and leverage advanced technologies such as containers, serverless computing, and more. We help Supercell to quickly develop, deploy, and scale their games to cope with varying numbers of gamers accessing the system throughout the course of the day.

AWS 119

AWS Cloud Games Serverless 119

All Things Distributed

JUNE 21, 2017

DynamoDB was the first service at AWS to use SSD storage. These high-throughput, low-latency requirements need caching, not as a consideration, but as a best practice. You just point your existing application at the DAX endpoint, and as a read-through/write-through cache, DAX seamlessly handles caching for you.

Speed 95

Speed Cache Latency AWS 95

The Morning Paper

JUNE 23, 2019

Coupled with stateless application servers to execute business logic and a database-like system to provide persistent storage, they form a core component of popular data center service archictectures. The network latency of fetching data over the network, even considering fast data center networks. Who knew! ;). From RInK to LInK.

Cache 79

Cache Latency Google Network 79

Percona

MAY 15, 2023

This operation is quite expensive but our database can run it in a few milliseconds or less, thanks to several optimizations that allow the node to execute most of them in memory with no or little access to mass storage. Of course, we are still talking about a low number of rows and running threads but the gain is there.

Traffic 144

Traffic Scalability Database Servers 144

All Things Distributed

NOVEMBER 19, 2010

Big news this week was of course the launch of Cluster GPU instances for Amazon EC2. Understanding Throughput-Oriented Architectures - background article in CACM on massively parallel and throughput vs latency oriented architectures. Driving Storage Costs Down for AWS Customers. Expanding the Cloud - The AWS Storage Gateway.

AWS 62

AWS Cloud Benchmarking Storage 62

The Morning Paper

MARCH 3, 2020

It takes you through the thinking processes and engineering practices behind the design of a key part of the control plane for AWS Elastic Block Storage (EBS): the Physalia database that stores configuration information. This work is latency critical, because volume IO is blocked until it is complete. This paper is a real joy to read.

Database 106

Database AWS Network Design 106

Percona

JULY 17, 2023

Alternatives to MongoDB Enterprise There are plenty of alternatives, but we’re not going to cover them all here, of course. Redis can handle a high volume of operations per second, making it useful for running applications that require low latency. Such organizations might seek a more straightforward alternative.

Open Source 98

Open Source Database Scalability Software 98

The Morning Paper

JUNE 13, 2019

Making queries to an inference engine has many of the same throughput, latency, and cost considerations as making queries to a datastore, and more and more applications are coming to depend on such queries. First off there still is a model of course (but then there are servers hiding behind a serverless abstraction too!). autoscaling).

Hardware 58

Hardware Latency Serverless Energy 58

The Morning Paper

JANUARY 26, 2020

Reverb also take a snapshot of the client’s local storage (e.g. If the server-side responder is also being replayed, then Reverb inserts a new request into the server-side log… When the response is generated, Reverb buffers it and uses a model of network latency to determine where to inject the response into the client-side log.

Programming 86

Programming Servers Network Latency 86

The Morning Paper

FEBRUARY 23, 2020

Our approach differs substantially by (1) providing economic incentives for data to be contributed and integrated into existing schemas, (2) offering a SQL interface instead of graph based approaches, (3) including the computational and storage infrastructure in the architectural vision. An embodiment for structured data for IoT.

Blockchain 94

Blockchain Internet Internet IoT 94

Dotcom-Montior

MAY 12, 2020

Websites are now more than just the storage and retrieval of information to present content to users. They now allow users to interact more with the company in the form of online forms, shopping carts, Content Management Systems (CMS), online courses, etc. Network latency. Network Latency. The list goes on and on.

Monitoring 63

Monitoring Entertainment Hardware Traffic 63

Highly Scalable

MAY 1, 2012

This approach often leads to heavyweight high-latency analytical processes and poor applicability to realtime use cases. The straightforward approaches for implementation of this system are: Log all events in a large storage like Hadoop and compute unique visitor periodically using heavy MapReduce jobs or whatever.

Analytics 191

Analytics Traffic Big Data Efficiency 191

n0derunner

SEPTEMBER 17, 2018

As with traditional storage, applications are writing to a shared storage environment which is necessary to support VM movement. It is the shared storage that often causes performance issues for data bases which are otherwise separated across nodes. The idea here is to simulate the two main storage workloads of a DB.

Performance Testing 40

Performance Testing Benchmarking Testing Performance 40

The Morning Paper

DECEMBER 8, 2019

Therefore any programming abstraction must be low latency and the kernel needs to be kept off the path of persistent data access as much as possible. Traditional pointers address a memory location (often virtual of course). This means that the overheads of system calls become much more noticeable.

Hardware 56

Hardware Programming Media Storage 56

The Morning Paper

SEPTEMBER 5, 2019

For response times (latencies) reporting a simple metric such as ‘average’ is next to useless. Instead we want to understand what’s happening at different latency percentiles (e.g A characteristic of latency distributions is that they have a long tail. Why do we need a new sketch? At the 50th percentile, an accuracy within 0.5%

Latency 52

Latency Metrics Open Source Java 52

SQL Performance

APRIL 8, 2019

Back on December 5, 2017, Microsoft announced that they were using AMD EPYC 7551 processors in their storage-optimized Lv2-Series virtual machines. Of course, if I were buying an AMD-based server for on-premises SQL Server use right now, I would try to get the newer, frequency-optimized AMD EPYC 7371 processor. Memory (GiB).

Azure 42

Azure Virtualization Storage Benchmarking 42

VoltDB

SEPTEMBER 10, 2024

For a company with 100 employees, multiply that over the course of the year, and you’re looking at upwards of 60,000 hours wasted. Typically, this involves using software and data virtualization tools to aggregate data from different databases, applications, and storage repositories.

IoT 52

IoT Architecture Scalability Analytics 52

O'Reilly

DECEMBER 2, 2019

For applications like communication between AVs, latency–how long it takes to get a response–is more likely to be a bigger limitation than raw bandwidth, and is subject to limits imposed by physics. There are impressive estimates for latency for 5G, but reality has a tendency to be harsh on such predictions.

Wireless 55

Wireless Serverless IoT Artificial Intelligence 55

Alex Russell

OCTOBER 22, 2017

It simulates a link with a 400ms RTT and 400-600Kbps of throughput (plus latency variability and simulated packet loss). Simulated packet loss and variable latency, however, can make benchmarking extremely difficult and slow. Our baseline, then, should probably trade lower throughput/higher-latency for packet loss.

Performance 74

Performance Network Benchmarking Mobile 74

The Morning Paper

MAY 28, 2019

Of course analog signalling comes with a whole bunch of challenges of its own, which is one of the reasons we tend to convert to digital. The resulting system can integrate seamlessly into a scikit-learn based development process, and dramatically reduces the total energy usage required for classification with very low latency.

Energy 52

Energy Hardware Efficiency Open Source 52

VoltDB

JUNE 26, 2024

This entails high-speed networks, real-time data platforms, scalable storage solutions, edge computing infrastructure, IoT devices, and advanced data processing capabilities. Both of these tactics can take a lot of time and require a lot of resources, which can pose a barrier to rapid intelligent manufacturing adoption.

IoT 52

IoT Analytics Artificial Intelligence Energy 52

Alex Russell

JULY 14, 2021

Here, native apps are doing work related to their core function; storage and tracking of user data are squarely within the four corners of the app's natural responsibilities. Fixing mobile won't be sufficient to unwind desktop's increasingly negative dark patterns, of course. But that's no reason to delay.

Google 134

Google Mobile Engineering Internet 134

The Agile Manager

JULY 9, 2010

Instead of producing large returns, of course, investment banking can produce large losses. This includes things like data storage, servers, e-mail, office productivity applications, virus protection, security and so forth. Of course, these complex properties don't make them "high value added" services. They're still utilities.

Retail 45

Retail Technology Technology Small Business 45

Alex Russell

APRIL 29, 2021

Chrome has missed several APIs for 3+ years: Storage Access API. For heavily latency-sensitive use-cases like WebXR, this is a critical component in delivering a good experience. It of course is personally frustrating to not be able to deliver these improvement for developers who wish to reach iOS users — which is all developers.

Media 145

Media Games Education Engineering 145

Brendan Gregg

SEPTEMBER 5, 2017

Here's some output from my zfsdist tool, in bcc/BPF, which measures ZFS latency as a histogram on Linux: # zfsdist. Tracing ZFS operation latency. Appliance manufacturers hire kernel engineers to develop custom features, including storage appliances. Hit Ctrl-C to end. ^C

Virtualization 75

Virtualization AWS Engineering Hardware 75

Highly Scalable

SEPTEMBER 18, 2012

Historically, NoSQL paid a lot of attention to tradeoffs between consistency, fault-tolerance and performance to serve geographically distributed systems, low-latency or highly available applications. Read/Write latency. Read/Write requests are processes with a minimal latency. Data Placement. Read/Write scalability.

Database 213

Database Latency C++ Scalability 213

High Scalability

JULY 20, 2018

A typical example of modern "microservices-inspired" Java application would function along these lines: Netflix : We observed during experimentation that RAM random read latencies were rarely higher than 1 microsecond whereas typical SSD random read speeds are between 100–500 microseconds.

Internet 121

Internet Internet Scalability Automotive 121

Alex Russell

MARCH 6, 2021

A then-representative $200USD device had 4-8 slow (in-order, low-cache) cores, ~2GiB of RAM, and relatively slow MLC NAND flash storage. Not as much as we'd like, of course, but the worldwide baseline has changed enormously. There are differences, of course, but not where it counts. The Moto G4 , for example.

Performance 127

Performance Network Mobile Metrics 127

Smashing Magazine

JANUARY 11, 2021

This might be very different for your company, of course, but that’s a close enough approximation of a majority of customers out there. Estimated Input Latency tells us if we are hitting that threshold, and ideally, it should be below 50ms. Thanks to Tim Kadlec, Henri Helvetica and Alex Russel for the pointers! Large preview ).

Performance 145

Performance Cache Media Metrics 145

Smashing Magazine

JANUARY 6, 2020

This might be very different for your company, of course, but that’s a close enough approximation of a majority of customers out there. Estimated Input Latency tells us if we are hitting that threshold, and ideally, it should be below 50ms. Thanks to Tim Kadlec, Henri Helvetica and Alex Russel for the pointers!

Performance 112

Performance Cache Servers Network 112

Smashing Magazine

JANUARY 7, 2019

Estimated Input Latency tells us if we are hitting that threshold, and ideally, it should be below 50ms. Of course, your data might show that your customers are not on these devices, but perhaps they simply don’t show up in your analytics because your service is inaccessible to them due to slow performance.

Performance 111

Performance Cache Network Metrics 111