<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Elasticsearch on Jorge Luis Betancourt</title><link>https://jorgelbg.me/tags/elasticsearch/</link><description>Recent content in Elasticsearch on Jorge Luis Betancourt</description><generator>Hugo</generator><language>en-us</language><lastBuildDate>Wed, 01 Jan 2020 11:46:14 -0500</lastBuildDate><atom:link href="https://jorgelbg.me/tags/elasticsearch/index.xml" rel="self" type="application/rss+xml"/><item><title>Better URL search with Elasticsearch</title><link>https://jorgelbg.me/2020/01/better-url-search-with-elasticsearch/</link><pubDate>Wed, 01 Jan 2020 11:46:14 -0500</pubDate><guid>https://jorgelbg.me/2020/01/better-url-search-with-elasticsearch/</guid><description>&lt;blockquote&gt;&#10;&lt;p&gt;🚀 This article was posted in the &lt;a href="https://tech.trivago.com/2020/02/11/better-url-search-with-elasticsearch/"&gt;trivago tech blog&lt;/a&gt;.&lt;/p&gt;&#10;&lt;/blockquote&gt;&#10;&lt;p&gt;At trivago, we generate a huge amount of logs and we have our &lt;a href="https://tech.trivago.com/2016/01/19/logstash_protobuf_codec/"&gt;own custom setup for shipping&#10;logs&lt;/a&gt; using mostly &lt;a href="https://developers.google.com/protocol-buffers"&gt;Protocol&#10;Buffers&lt;/a&gt;. Eventually we end up with some fields in ES&#10;that contain partial (or full) URLs. For instance in our specific case we store the &lt;a href="https://en.wikipedia.org/wiki/URL#Syntax"&gt;query&#10;component&lt;/a&gt; of the URL in a field called &lt;code&gt;query&lt;/code&gt; and the&#10;&lt;a href="https://en.wikipedia.org/wiki/URL#Syntax"&gt;path component&lt;/a&gt; in a field named &lt;code&gt;url_path&lt;/code&gt;. Sample values&#10;for these fields could be:&lt;/p&gt;</description></item><item><title>Useful alerts with Elastic Watcher &amp; Machine Learning</title><link>https://jorgelbg.me/2019/03/useful-alerts-with-elastic-watcher-machine-learning/</link><pubDate>Fri, 15 Mar 2019 13:00:00 +0100</pubDate><guid>https://jorgelbg.me/2019/03/useful-alerts-with-elastic-watcher-machine-learning/</guid><description>&lt;p&gt;Alerts should be meaningful, simply getting a notification about something&#10;happening is not enough. A &lt;a href="https://blog.danslimmon.com/2017/10/02/what-makes-a-good-alert/"&gt;good alert&lt;/a&gt;&#10;should be &lt;strong&gt;actionable&lt;/strong&gt; and &lt;strong&gt;investigable&lt;/strong&gt;. For me, this means getting a&#10;answer to the following questions:&lt;/p&gt;&#10;&lt;h3 id="what-happened"&gt;What happened?&lt;/h3&gt;&#10;&lt;p&gt;Was the traffic to the Web site unusually low? Did the number of errors&#10;increase? This is the type of information that the alert should contain in a&#10;clear and visible way.&lt;/p&gt;&#10;&lt;h3 id="why-an-alert-triggered"&gt;Why an alert triggered?&lt;/h3&gt;&#10;&lt;p&gt;What does it mean if the traffic is &lt;em&gt;unusually low&lt;/em&gt;, would a 5% qualify for&#10;this category? What about 10%? How much are too many errors? What &lt;em&gt;type&lt;/em&gt; of&#10;errors has increased?&lt;/p&gt;</description></item><item><title>Optimize Grafana dashboards with Elasticsearch index aliases</title><link>https://jorgelbg.me/2018/12/optimize-grafana-dashboards-with-elasticsearch-index-aliases/</link><pubDate>Thu, 20 Dec 2018 00:00:00 +0100</pubDate><guid>https://jorgelbg.me/2018/12/optimize-grafana-dashboards-with-elasticsearch-index-aliases/</guid><description>&lt;p&gt;Grafana is a very popular opensource dashboarding solution. Provides support&#10;for a long list of storage solutions, including Elasticsearch.&#10;Unfortunately, the ES support is not at the same level as the one you get for&#10;InfluxDB, for instance. Still, Grafana allows combining in the same&#10;dashboard different data sources. It is possible to have a panel fetching data&#10;from ES and a different panel fetching data from InfluxDB.&lt;/p&gt;&#10;&lt;h2 id="grafana--elasticsearch"&gt;Grafana ❤️ Elasticsearch&lt;/h2&gt;&#10;&lt;p&gt;Grafana provides stellar support for &lt;a href="https://www.influxdata.com/"&gt;InfluxDB&lt;/a&gt;&#10;&amp;amp; &lt;a href="http://docs.grafana.org/features/datasources/prometheus/"&gt;Prometheus&lt;/a&gt;,&#10;among others. This means that you get, query autocompletion for fields, values,&#10;etc. When you select ES as a data source for a panel, the features are a bit less&#10;polished. You are greet by a &amp;ldquo;Lucene query&amp;rdquo; input and some more options.&#10;Depending on the metric (aggregation), Grafana also provides some help&#10;for field name selection when using ES as a data source.&lt;/p&gt;</description></item></channel></rss>