VictoriaMetrics Observability Blog

Author
Category

Filter: Time Series

The Life of a Metric

Follow a single metric through its whole life inside VictoriaMetrics: born as a counter in your process, exported over the wire, given an identity and an inverted-index entry, written through an LSM tree, compressed into columnar blocks on disk, read back by a rate() query, downsampled as it ages, and finally deleted when retention runs out.

VictoriaMetrics Components: Getting Started

VictoriaMetrics is a fast, scalable monitoring system made of modular components like vminsert, vmstorage, and vmselect. It supports both single-node and clustered setups, along with tools for backup, restore, alerting, access control, and data migration. Data can be ingested, stored, queried, backed up, and restored with high performance and minimal resource use.

Prometheus Alerting 101: Rules, Recording Rules, and Alertmanager

Recording rules can transform complex PromQL queries into pre-computed metrics, eliminating repeated calculation overhead. When organized in groups, these alerting and recording rules can be precisely scheduled to run at specific times like 01:30, 02:30, and even execute in parallel for maximum efficiency.

Inside vmselect: The Query Processing Engine of VictoriaMetrics

The article explains vmselect’s core functionalities: concurrent request processing, query parsing and validation, data fetching and merging from vmstorage nodes, and memory-efficient result caching.

How vmstorage Handles Query Requests From vmselect

When vmselect asks vmstorage for data, there’s actually quite a bit happening behind the scenes. This article breaks down the process of how vmstorage locates and retrieves your metrics, from finding the right TSIDs to sending back the actual data points.

How vmstorage's IndexDB Works

IndexDB acts as vmstorage’s memory - it remembers which numbers (TSIDs) belong to which metrics, making sure your queries get answered fast. This article walks through how this system works, from the way it organizes data to how it keeps track of millions of timeseries.

How vmstorage Processes Data: Retention, Merging, Deduplication,...

vmstorage takes data from vminsert through a concurrency limiter, creates TSIDs for each row, and puts them in memory buffers. Every few seconds, it moves data to storage parts on disk. The system merges parts, removes duplicates, and cleans old data. This turns raw metrics into data users can search.

The Rise of Open Source Time Series Databases

Time series databases are essential tools in any software engineer’s toolbelt. Their development has been shaped by user needs and countless open source contributors, leading to the healthy ecosystem of options we see today. In this article, you’ll see how time series databases came about, and why so many are open source.

VictoriaMetrics Cloud reduces monitoring costs by 5x

We’re happy to announce VictoriaMetrics Cloud, a hosted monitoring platform and managed service for metrics that allows organizations to monitor and store large amounts of time-series data, without having to run the underlying infrastructure.

What’s new in VictoriaMetrics 2021?

With more than 20 new releases of VictoriaMetrics published during 2021, a features roundup seemed appropriate. This blog walks you through the key VM features released in 2021.