# 7.3.2 и потеря мастера

**URL:** <https://discuss.elastic.co/t/7-3-2/201228>\
**Category:** Вопросы на русском языке\
**Created:** [September 26, 2019, 12:24pm UTC](https://discuss.elastic.co/t/7-3-2/201228 "2019-09-26T12:24:20Z")\
**Posts on this page:** 1\
**Showing post:** 13

<div class="post-metadata">

**Author:** ![Denis\_Lamanov](https://sea2.discourse-cdn.com/elastic/user_avatar/discuss.elastic.co/denis_lamanov/32/13111_2.png) [@Denis\_Lamanov](https://discuss.elastic.co/u/Denis_Lamanov)\
**Post date:** [October 12, 2019, 4:09pm UTC](https://discuss.elastic.co/t/7-3-2/201228/13 "2019-10-12T16:09:24Z")

</div>

Судя по логам т.к. у нас не SSD диски, то обновление мета-информации о каждоом индексе(которых у нас больше чем 1000 на ноду) на диске иногда превышает `cluster.publish.timeout` ([Rolling upgrade problem from 6.8 to 7.1.1](https://discuss.elastic.co/t/rolling-upgrade-problem-from-6-8-to-7-1-1/186571/14))  
Собственно этот механизм, судя по документации, ввели в 7-ке  
[https://www.elastic.co/guide/en/elasticsearch/reference/7.4/cluster-state-publishing.html](https://www.elastic.co/guide/en/elasticsearch/reference/7.4/cluster-state-publishing.html)

---

_[View the full topic](https://discuss.elastic.co/t/7-3-2/201228)._