Confluence Data Center でコンテンツ インデックスをゼロから再構築する方法
プラットフォームについて: Server と Data Center のみ - この記事は、サーバーおよびデータセンター プラットフォームのアトラシアン製品にのみ適用されます。
This KB pertains to Confluence Data Center (cluster) only. If you are running Confluence Server (single-server), please see How to Rebuild the Content Indexes From Scratch on Confluence Server instead.
The Confluence index may become out of date or corrupted due to a variety of reasons (see Searching and Indexing Troubleshooting). When this occurs, some pages may not show up in search results as expected. The purpose of this KB is to detail the steps on how to rebuild the index from scratch across the entire Confluence Data Center cluster while minimizing downtime.
Rebuilding the index can sometimes take a while as it is re-indexing all the content in your site. Please be aware that while a node is rebuilding the index, users will be unable to search Confluence on that node. As such, you may wish to temporarily remove the node from the rotation in your load balancer so that users will not access it. Additionally, please consider performing your re-index after hours.
Solution: Confluence 7.7 and later
From Confluence 7.7 rebuilding the index and propagating the new index to all nodes in your cluster can be done directly through the Confluence UI.
This method requires no downtime, and you can also choose to remove the node performing the reindex from your load balancer, to further minimise any performance impact on users. Confluence will continue to use the existing index until the new index has been successfully rebuilt and propagated to each node in the cluster.
See Content Index Administration for more information.
Solution: Confluence 5.9 to 7.6
Rebuild the index with downtime
Refer to the following KB article to rebuild the content index if you can afford some downtime on your instance:
How to manually rebuild content index from scratch on Confluence Data Center with downtime.
Rebuild the index without downtime
Refer to the following KB article to rebuild the content index without any downtime: