How we knew it was time to leave the cloud

The challenges of scaling GitLab's infrastructure have led to a shift from using cloud-based CephFS storage to dedicated hardware. Initially, CephFS was chosen for its scalability and capacity, but its performance on cloud infrastructure was hindered by shared resources and IO latency, resulting in system bottlenecks and high costs as performance demands increased. This prompted a move to bare metal solutions, which, despite upfront costs, offer more reliable and economical performance for GitLab's needs. To better understand and address these challenges, GitLab has invested in building an observable system that uses Prometheus for gathering and visualizing system metrics, enabling the team to proactively identify and resolve infrastructure issues. This approach is part of a broader strategy to enhance monitoring capabilities, including the development of a gitlab-monitor Prometheus exporter, to provide deeper insights into system performance and trends.

No tracked trend matches for this post yet.

Use this post, company, and trend context to find content marketing opportunities, perform competitive analysis, or address product feature gaps via the Plushcap MCP server or the Plushcap API.