<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Api on Kuldeep Pisda</title><link>https://kdpisda.in/tag/api/</link><description>Recent content in Api on Kuldeep Pisda</description><generator>Hugo</generator><language>en-us</language><lastBuildDate>Fri, 22 May 2026 10:00:00 +0530</lastBuildDate><atom:link href="https://kdpisda.in/tag/api/index.xml" rel="self" type="application/rss+xml"/><item><title>Retrying Third-Party APIs the Right Way: Honor the Headers</title><link>https://kdpisda.in/retrying-third-party-apis-honor-rate-limit-headers/</link><pubDate>Fri, 22 May 2026 10:00:00 +0530</pubDate><guid>https://kdpisda.in/retrying-third-party-apis-honor-rate-limit-headers/</guid><description>&lt;p&gt;You call a third-party API. It hands you back a &lt;code&gt;429 Too Many Requests&lt;/code&gt;. What do you do?&lt;/p&gt;
&lt;p&gt;If your reflex is &amp;ldquo;wait a second, then two, then four, then hope&amp;rdquo; — same. That was my reflex too. Exponential backoff is the pattern everyone reaches for, it&amp;rsquo;s in every resilience blog post, and it feels responsible. It&amp;rsquo;s also, in a lot of cases, just guessing. The API you&amp;rsquo;re hammering usually tells you &lt;em&gt;exactly&lt;/em&gt; how long to wait. You&amp;rsquo;re choosing to ignore it and roll dice instead.&lt;/p&gt;</description></item></channel></rss>