--- title: Why you should stop using ARBITERS in MongoDB | daily.dev description: ARBITER nodes in MongoDB solve a voting problem but never a durability problem, so Percona recommends dropping them from production. In one case, losing a data... canonical: https://daily.dev/posts/why-you-should-stop-using-arbiters-in-mongodb-3xehjgqjz twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: Why you should stop using ARBITERS in MongoDB | daily.dev og:description: ARBITER nodes in MongoDB solve a voting problem but never a durability problem, so Percona recommends dropping them from production. In one case, losing a data... og:url: https://daily.dev/posts/why-you-should-stop-using-arbiters-in-mongodb-3xehjgqjz og:image: https://api.daily.dev/og/posts/3XEhjGQJZ.png og:image:alt: Why you should stop using ARBITERS in MongoDB og:image:width: 1200 og:image:height: 630 og:locale: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # Why you should stop using ARBITERS in MongoDB **[Percona Blog](https://daily.dev/sources/percona)** · 7 min read · 0 upvotes · 0 comments ## Summary ARBITER nodes in MongoDB solve a voting problem but never a durability problem, so Percona recommends dropping them from production. In one case, losing a data center dropped write throughput from 3-4k ops/s per shard to just 50-75 ops/s, because majority write concern needed four data-bearing acknowledgments the arbiter couldn't provide. In another, an hour of acknowledged writes on a temporary primary got rolled back after a node outage, since the arbiter held no data to preserve them. MongoDB 5.3 disabled multiple arbiters by default, and Atlas doesn't support arbiters at all. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Questions this post answers ### Why did my MongoDB cluster's write throughput collapse after a data center failure even though enough nodes were still reachable including an arbiter? Write throughput collapses because majority write concern counts only data-bearing acknowledgments, and an arbiter never counts toward that majority even though it votes in elections. In one documented case, a sharded cluster with three data-bearing nodes per shard plus an arbiter saw throughput fall from 3-4k ops/s to 50-75 ops/s per shard after losing a data center, because majority still required four data-bearing acknowledgments. _Anyone troubleshooting MongoDB replica set performance can dig into cases like this one on daily.dev._ ### Can MongoDB roll back acknowledged writes after a primary failover? Yes, writes acknowledged by a temporary primary can be rolled back if that node goes down before the original primary catches up. In one case, an hour of writes acknowledged with w:1 on a secondary promoted to primary were rolled back to bson files under the rollback directory in dbPath once the original primary, which never received them, resumed as primary. _Teams designing MongoDB write concerns for durability can track real failure cases like this on daily.dev._ ### Does MongoDB still support multiple arbiters in a replica set? No, starting in MongoDB 5.3, support for multiple arbiters in a replica set is disabled by default because multiple arbiters prevent reliable use of the majority write concern and can let a replica set accept writes without sufficient secondaries for data replication. Arbiters also block shard key changes via transactions and multi-shard transactions on shards containing an arbiter. _Developers planning MongoDB version upgrades can follow changes like this on daily.dev._ ## Similar posts on daily.dev - [How We Built an Automated, Production-Ready Sharded MongoDB Infrastructure](https://daily.dev/posts/how-we-built-an-automated-production-ready-sharded-mongodb-infrastructure-n1vi9g1lo) · Trendyol Tech · 1 upvotes · 0 comments - [Multi-Cluster databases on Kubernetes: Architecture and deployment](https://daily.dev/posts/multi-cluster-databases-on-kubernetes-architecture-and-deployment-yats0aq1j) · CNCF · 1 upvotes · 0 comments - [AI Is Changing What Customers Need From a Database. MongoDB 8.3 Is Built for It](https://daily.dev/posts/ai-is-changing-what-customers-need-from-a-database-mongodb-8-3-is-built-for-it-clhjgtncp) · MongoDB · 1 upvotes · 0 comments - [Why Percona Backup for MongoDB Is the Right Choice for Production Backups](https://daily.dev/posts/why-percona-backup-for-mongodb-is-the-right-choice-for-production-backups-aovk7aqvo) · Percona Blog · 0 upvotes · 0 comments --- Tags: [#database](https://daily.dev/tags/database), [#mongodb](https://daily.dev/tags/mongodb), [#distributed-systems](https://daily.dev/tags/distributed-systems) [View this post on daily.dev](https://daily.dev/posts/why-you-should-stop-using-arbiters-in-mongodb-3xehjgqjz) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@type":"TechArticle","headline":"Why you should stop using ARBITERS in MongoDB","url":"https://daily.dev/posts/why-you-should-stop-using-arbiters-in-mongodb-3xehjgqjz","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/why-you-should-stop-using-arbiters-in-mongodb-3xehjgqjz"},"datePublished":"2026-10-28T23:53:45.857Z","dateModified":"2026-10-29T00:04:02.389Z","description":"ARBITER nodes in MongoDB solve a voting problem but never a durability problem, so Percona recommends dropping them from production. In one case, losing a data...","image":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/b37b299bdb39718d2b13f1c9728d12d7?_a=AQAEuop","thumbnailUrl":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/b37b299bdb39718d2b13f1c9728d12d7?_a=AQAEuop","isAccessibleForFree":true,"articleSection":"Percona Blog","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"Percona Blog","logo":"https://media.daily.dev/image/upload/s--Lb3XkBh3--/f_auto,q_auto/v1780213412/logos/percona?_a=BAMAMiWQ0","url":"https://daily.dev/sources/percona"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/why-you-should-stop-using-arbiters-in-mongodb-3xehjgqjz","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":0},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"database,mongodb,distributed-systems","timeRequired":"PT7M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"Percona Blog","item":"https://daily.dev/sources/percona"},{"@type":"ListItem","position":3,"name":"Why you should stop using ARBITERS in MongoDB"}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/why-you-should-stop-using-arbiters-in-mongodb-3xehjgqjz#faq","mainEntity":[{"@type":"Question","name":"Why did my MongoDB cluster's write throughput collapse after a data center failure even though enough nodes were still reachable including an arbiter?","acceptedAnswer":{"@type":"Answer","text":"Write throughput collapses because majority write concern counts only data-bearing acknowledgments, and an arbiter never counts toward that majority even though it votes in elections. In one documented case, a sharded cluster with three data-bearing nodes per shard plus an arbiter saw throughput fall from 3-4k ops/s to 50-75 ops/s per shard after losing a data center, because majority still required four data-bearing acknowledgments. Anyone troubleshooting MongoDB replica set performance can dig into cases like this one on daily.dev."}},{"@type":"Question","name":"Can MongoDB roll back acknowledged writes after a primary failover?","acceptedAnswer":{"@type":"Answer","text":"Yes, writes acknowledged by a temporary primary can be rolled back if that node goes down before the original primary catches up. In one case, an hour of writes acknowledged with w:1 on a secondary promoted to primary were rolled back to bson files under the rollback directory in dbPath once the original primary, which never received them, resumed as primary. Teams designing MongoDB write concerns for durability can track real failure cases like this on daily.dev."}},{"@type":"Question","name":"Does MongoDB still support multiple arbiters in a replica set?","acceptedAnswer":{"@type":"Answer","text":"No, starting in MongoDB 5.3, support for multiple arbiters in a replica set is disabled by default because multiple arbiters prevent reliable use of the majority write concern and can let a replica set accept writes without sufficient secondaries for data replication. Arbiters also block shard key changes via transactions and multi-shard transactions on shards containing an arbiter. Developers planning MongoDB version upgrades can follow changes like this on daily.dev."}}]} ```