# About storage server's recovery

**URL:** <https://forums.foundationdb.org/t/about-storage-servers-recovery/495>\
**Category:** FoundationDB Core\
**Created:** [June 7, 2018, 10:59am UTC](https://forums.foundationdb.org/t/about-storage-servers-recovery/495 "2018-06-07T10:59:00Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![coder0718](https://sea1.discourse-cdn.com/foundationdb/user_avatar/forums.foundationdb.org/coder0718/32/87_2.png) [@coder0718](https://forums.foundationdb.org/u/coder0718)\
**Post date:** [June 7, 2018, 10:59am UTC](https://forums.foundationdb.org/t/about-storage-servers-recovery/495/1 "2018-06-07T10:59:00Z")

</div>

when storage server reboot, how does it know which logs to recover from transaction log?

 ![image](https://global.discourse-cdn.com/foundationdb/original/1X/6253b6c7154b1a487afacf3fa14ab31ad5ba66ae.png)

---

<div class="post-metadata">

**Author:** ![Evan](https://sea1.discourse-cdn.com/foundationdb/user_avatar/forums.foundationdb.org/evan/32/104_2.png) [@Evan](https://forums.foundationdb.org/u/Evan)\
**Post date:** [June 8, 2018, 6:44am UTC](https://forums.foundationdb.org/t/about-storage-servers-recovery/495/2 "2018-06-08T06:44:22Z")

</div>

Every storage server has an associated tag in the log system. Each tag is essentially a queue of mutations needed by the associated storage server. The log system is designed so that one log server will have all the data any given tag. If that log server has failed, the storage server must ask all the other log servers for the data, and merge the results.