How Salesforce Eliminated Single-Region Risk and Reduced Downtime Blast Radius at 4B Metrics/Min

Summary

Salesforce redesigned its internal observability platform, Argus, to eliminate single-region failure risks while ingesting approximately 4 billion metrics per minute. Operating out of a single AWS region previously created a global blast radius and generated high cross-region data transfer costs. To mitigate these risks without the prohibitive expense of full multi-region replication, the engineering team implemented a geo-local architecture that processes and stores telemetry closer to its origin across production geographies. A newly introduced federation query layer relies on Elasticsearch mappings to route queries selectively to relevant regional clusters, avoiding broad fan-out. The platform also adopted metadata caching for wildcard queries and implemented HTTP 206 partial-response handling to maintain visibility when individual regions experience downtime.

Context

Salesforce operates Argus, an internal observability platform ingesting approximately 4 billion metrics per minute. Originally deployed in a single AWS region, the platform faced major availability risks because an outage in that single region would take down observability globally. Furthermore, routing all metrics across data centers and regions into one central location incurred high data transfer costs and network latency, while full multi-region replication was cost-prohibitive.

Approach / What changed

The team transitioned Argus to a geo-local architecture where metrics are processed and stored closer to their origin across production geographies using OpenTSDB and HBase. To query distributed data without full replication, a federation query layer was introduced that uses Elasticsearch mappings of metric scopes to selectively route queries to relevant regions instead of broadcasting. The system handles partial availability by returning HTTP 206 partial-response semantics with UI indicators, and accelerates wildcard query resolution using metadata caching.

Takeaways

• Argus adopted a geo-local model where telemetry is processed and stored near its origin across multiple geographies, avoiding the infrastructure costs of full multi-region data replication.
• A federation query layer uses Elasticsearch mappings to map metric scopes to regions, routing queries only to relevant geographies to prevent unnecessary cross-region fan-out and latency.
• To handle partial regional outages gracefully during distributed queries, Argus returns HTTP 206 partial responses and surfaces UI indicators instead of failing silently.

Related reading

• Salesforce · Aug 11, 2026

  How Standardizing Product Telemetry Reduced Time to Insight by 97%

  Salesforce faced scaling bottlenecks when individual engineering teams maintained fragmented, custom telemetry pipelines requiring manual data transformations and dashboard creation every release. To resolve these silos, the engineering organization established the Product Data Platform (PDP) using a standardized custom schema built atop Monitoring Cloud infrastructure. This framework mandates core telemetry fields for consistent metric generation while offering governed optional attributes for product-specific flexibility. An AI-driven Model Context Protocol tool assists developers by recommending compliant instrumentation patterns directly within their workflows. Today, the unified platform processes 45 billion rows of data daily across 19,000 distinct events and 2,000 product features, reducing dashboard refresh latency by 97 percent from one month to daily updates.

  Scott Nyberg

• Grab · May 1, 2017

  Scaling Like a Boss with Presto

  Grab experienced severe performance degradation, long queue times, and connection timeouts on its Amazon Redshift analytics cluster as user concurrency and reporting workloads expanded. Although an initial Amazon S3 data lake decoupled storage from compute, business users required standard SQL interfaces rather than Spark data pipelines. The team deployed Presto clusters on AWS EMR, switching their storage format from AVRO to Parquet to support ANSI SQL querying directly against S3. Utilizing a shared Hive metastore on Amazon RDS allowed Grab to adopt a shared-data multi-cluster architecture that isolated distinct workloads across dedicated compute clusters. This setup enabled rapid cluster scaling, streamlined failover, and matched Redshift performance on partitioned time-range queries.

  Aneesh Chandra

• Salesforce · Jul 23, 2026

  How AI Rebuilt Salesforce’s Decades-Old Localization Pipeline

  Salesforce needed to deliver major releases across 34 languages as product localization volume surged by more than 35% under fixed budgets and unchanged delivery windows. To address these scaling constraints, the engineering team replaced their legacy translation workflows with an LLM-powered orchestration pipeline integrating prompt engineering, context engineering, and multi-stage validation. The architecture transforms existing style guides, glossaries, and terminology databases into structured context, routing interface strings through specialized prompt stages—up to 85 stages for some languages—to enforce product context, grammar, and branding rules. Combined with an automated validation loop spanning AI editing, AI validation, and human review, this system reduced localization costs by 50 to 90% while accelerating turnaround times across enterprise releases.

  Scott Nyberg

• Salesforce · Jul 30, 2026

  How Salesforce Built an Agentic Engineering Enablement Strategy for Thousands of Software Engineers

  Salesforce addressed the challenge of scaling agentic engineering across thousands of software developers by focusing on organizational learning rather than tooling. While individual engineers initially developed disparate, uncoordinated workflows, the Technology, People, Innovation, and Learning (TPIL) team unified these approaches into the Proficiency Level (PL) Framework. The framework maps developer growth across four mindset-driven stages: AI-Assisted, AI-Validating, AI-Orchestrating, and AI-Native. To facilitate progress across these stages, TPIL introduced global AI camps, weekly working sessions, and manager coaching guides. Rather than tracking course completions or telemetry scores, Salesforce measured transformation through behavioral indicators, such as shifts toward validating agent output, designing multi-agent workflows, and asking higher-level questions regarding human judgment.

  Scott Nyberg