# DGO client does not handle replica failure since it does not have a proper connection pool

**URL:** <https://discuss.dgraph.io/t/dgo-client-does-not-handle-replica-failure-since-it-does-not-have-a-proper-connection-pool/18226>\
**Category:** Dgraph Clients\
**Tags:** dgo\
**Created:** [February 1, 2023, 4:54pm UTC](https://discuss.dgraph.io/t/dgo-client-does-not-handle-replica-failure-since-it-does-not-have-a-proper-connection-pool/18226 "2023-02-01T16:54:50Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![alexeevaa](https://yyz1.discourse-cdn.com/flex007/user_avatar/discuss.dgraph.io/alexeevaa/32/8341_2.png) [@alexeevaa](https://discuss.dgraph.io/u/alexeevaa)\
**Post date:** [February 1, 2023, 4:54pm UTC](https://discuss.dgraph.io/t/dgo-client-does-not-handle-replica-failure-since-it-does-not-have-a-proper-connection-pool/18226/1 "2023-02-01T16:54:50Z")

</div>

Once replica fails, its grpc connection is still present in a makeshift pool of DGO client. This leads to a stream of unnecessary errors while cluster is actually available. This is especially important in case you build a lambda server, which is shared between all alpha nodes.