Shreyansh Sancheti
SDE-II @IBM ISDL| Customer focus | Open source day GHC’23 mentor | Ceph Storage Developer | Cybersecurity Enthusiast | Seeking Dynamic Challenges
Actions
As an accomplished Software Engineer with extensive expertise in Ceph-based storage solutions, I possess a strong foundation in software engineering. I have showcased my programming skills in C++ and Python, which have been instrumental in my contributions to Ceph and various projects. My curiosity and drive to solve real-world problems are evident through my research projects, published articles, and a self-created CTF challenge in cybersecurity. I love to dance, sing and to watch anime.
Links
Performance evaluation of Ceph Bluestore RocksDB Compression: Gain space, lose nothing*
This presentation sheds light on an essential tuning enhancement in Ceph, specifically Bluestore RocksDB compression, now enabled by default in versions Reef and above. Through this presentation, our aim is to familiarize the broader audience with rocksDB compression’s significance, advantages, and usage in Ceph, particularly in hybrid
deployments where NVMe drives store WAL/DB data.
By exercising this optimization, we achieve nearly identical performance while considerably reducing space utilization for storing metadata.
In addition to explaining rocksDB compression, we will delve into the testing workflow employed during the performance evaluation, we emphasize on the methodologies deployed and the performance of Ceph under different sets of workloads with and without fault insertion.
Intelligent Tiering for RGW
Ceph can currently provide Tiering(Storage Classes) using different pools in the Ceph on-prem cluster and can also do cloud transitions(AWS, s3 compatible endpoints). The movement of data through different storage classes can be achieved using S3 lifecycle rules based on the age of the files. Intelligent tiering for RGW (Rados Gateway) is a feature in Ceph storage that automatically moves data between storage tiers based on policies. This feature optimises storage costs and performance by moving frequently accessed data to faster storage tiers and less frequently accessed data to lower-cost, higher-capacity tiers. The intelligent tiering policy is customizable, allowing users to define criteria for data migration, such as access frequency, time, and size. This feature is particularly useful for large-scale object storage deployments, where optimizing storage costs and performance is crucial. The first step is to extend the existing cloud transition to include features like read-through, restore API(like glacier in s3 protocol)
Enhance Read performance for RGW using local read.
Ceph's OSD support different flags for Read Operation like default, local and balanced. "Local Read" feature enables faster data retrieval by allowing clients to read data directly from local OSD, reducing latency and improving performance. When enabled, RGW directs read requests to the most suitable local OSD, with help of crush location set by the user. This feature is particularly beneficial for applications requiring fast data access, such as content delivery networks (CDNs), big data analytics, and AI/ML workloads. This can be easily configured for Stretch clusters , Kubernetes cluster etc.
Optimizing Ceph RGW: Tuning Strategies for Enhanced Performance
In the dynamic world of storage optimization, fine-tuning Ceph RGW (RADOS Gateway) configurations can significantly impact performance and efficiency. However, navigating the multitude of tuning options can be daunting. This lightning talk delves into various RGW tuning strategies, offering insights into both the benefits and potential pitfalls. From adjusting thread pools to optimizing objector and frontend parameters, we explore practical use cases, scenarios, and considerations for standalone and multisite clusters. Drawing from real-world experiences and community insights, we highlight the importance of testing, caution against over-tuning, and emphasize the need for informed decision-making. Attendees will gain actionable insights into RGW tuning, enabling them to optimize their Ceph deployments effectively.
What my audience will learn from my presentation ?
1. Key RGW tuning parameters and their impact on performance.
2. Practical use cases and scenarios for optimizing standalone and multisite clusters.
3. Considerations for balancing performance, latency, and memory usage.
4. The importance of testing and informed decision-making in RGW optimization.
Shreyansh Sancheti
SDE-II @IBM ISDL| Customer focus | Open source day GHC’23 mentor | Ceph Storage Developer | Cybersecurity Enthusiast | Seeking Dynamic Challenges
Links
Actions
Please note that Sessionize is not responsible for the accuracy or validity of the data provided by speakers. If you suspect this profile to be fake or spam, please let us know.
Jump to top