Zach Seils

I wanted to seek the advice of my friends, colleagues and peers regarding AWS EKS. it seems that the default settings are to build it with the API public to the internet. This feels like a very bad Idea to me, but on the plus side, it means my helm charts are very easily deployed by git hub runners. So my question, what do people do regarding EKS API access, and if people keep it private only, how do they then connect from services outside of AWS like GitHub actions? Thanks in advance... ** UPDATE ** Just to close the loop on this, and many thanks for all of your thoughts and suggestions. I've implemented a hosted Github runner inside our VPC which is running perfectly. Thanks :-)

5

16 Comments

Take a look at ☁️ Emanuel Russo latest blog: it would be helpful for software engineers, specially in this season 😁

7

Manage access of your content in S3 using OTIO!

4

Today, AWS Config expands support for 124 resource types across 29 AWS Regions. If you have enabled AWS Config recording for these resource types, you will benefit from more comprehensive coverage across additional AWS Regions.

14

A new post for my blog this morning, on the great new paper from the Amazon MemoryDB team: https://lnkd.in/eeKYfzBW This paper is a great reminder that distributed systems aren't just about scalability. Availability, durability, consistency, performance, and consistent performance are all equally important. Distribution makes them easier to achieve. Read the team's paper here: https://lnkd.in/eWxuuSTA

415

2 Comments

We are pleased to announce that you can now achieve up to an aggregate 1.3 million I/O operations per second (IOPS) and 12 GB/s throughput per OCI compute instance with the OCI Block Volume Service. In addition, you can also attach up to 32 Ultra High Performance (UHP) volumes to a single OCI compute instance. This service update is particularly useful for high I/O workloads, such as OLTP, AI/ML, data warehousing and analytics, databases with large data sets and extensive complex processing. https://lnkd.in/eHErX5mC

2

This is a big deal for using AI in production - Claude 3.5 Sonnet and 3 Haiku are now available in more regions. us-west-2 tends to get new AI stuff first, and it is great to see these 2 models now more geographically available. https://lnkd.in/gvQkMQbX

17

C4 VMs now GA: Unmatched Performance and Control for Enterprise Workloads. Up to 20% better price-performance for general purpose workloads & 45% better price-performance for CPU-based inference.

1

Do you need an Amazon EC2 instance with a truly stupendous amount of memory? Check out my newest blog post to learn about the new U7i high memory instances with 12 TiB to 32 TiB of DDR5 memory and 896 vCPUs, available today in four #AWS Regions. https://lnkd.in/gVgxJ-bF

771

65 Comments

We built MSK Replicator to make it easy for customers to operate multi-region #Kafka setups for disaster recovery use cases. With MSK Replicator, customers get fully managed networking for cross-region Amazon Managed Streaming for Apache Kafka (MSK) setups. A key customer ask, post-launch, has been to use MSK Replicator without having to make any code changes to their applications. Customers want to have the same topic name across both their primary and their secondary clusters. This comes however, at a risk of potential circular replication of data across both clusters. At AWS we’ve been building highly resilient services for many years. And getting the resiliency just right for MSK Replicator has been a key focus for us. Services like Amazon S3 have been operating multi-region for years and we’ve taken their learnings and are now releasing a new capability to allow customers to have the same topic names but with robust protection against circular replication. This means customers can expand their Kafka setups across regions without any code changes. #ApacheKafka #MSK #AWS #Day1 #releaseday

27

1 Comment

Build your transcoder with FFMPEG 7.0.1

30

Indeed, OpenTofu and Valkey, being integral parts of The Linux Foundation, offer a robust open source framework that is well-positioned for long-term sustainability and resilience. OpenTofu, emerging as a community-driven alternative to Terraform Labs, promises an open and modular approach to infrastructure as code, backed by a strong commitment from over 140 organizations and 600 individuals. Similarly, Valkey, the open source successor to Redis, ensures continued collaborative development of its high-performance in-memory data store, maintaining the project under the open source BSD 3-clause license and supported by industry giants. #opensource #database #community #developers

12

AWS deprecating a few services. "S3 Select, CloudSearch, Cloud9, SimpleDB, Forecast, Data Pipeline, and CodeCommit" The one-stop-cloud-shop was never a thing. At least for me. https://lnkd.in/dxc6FrTy #aws #devops #architecture

9

1 Comment

Adding support for new protocols in fixed pipeline packet processing is usually a nightmare unless the pipeline has enough flexibility like we have in Express5! When we were tasked to support SRv6 natively in Express5, our packet processing architect, Nancy Shaw, and the team came up with a novel way to implement the protocol natively using the in-built flexibility and some additional enhancements. ✔ Variable header (up to 256B) to the pipeline ✔ Flexible parser/rewrite ✔ Shared memory with fungible partitions to store uSID arrays, ✔ Recirculation to look deeper into the packet ✔ Novel patented technique to do two lookups (current and next uSID) without performance penalty ✔ .... The resulting implementation is superior in both area and power, handling SRv6 packets with up to 48 uSIDs at 5BPPs. Read the article below to gain more insights! https://lnkd.in/eDZJDwsF #segmentrouting #ipv6 #ipv4 #srv6 #ietf

163

5 Comments

A new blog post just been released! This is a new benchmark about fan-out on AWS with Dynamo Stream, EventBridge, and SQS, in compare with Dynamo stream, Lambda, SNS and SQS. The more serverless the quicker? https://lnkd.in/dgh5d_W2

6