# Crash régulier
**URL:** https://discuss.elastic.co/t/crash-regulier/163795
**Category:** Discussions en français
**Created:** [January 10, 2019, 6:02pm UTC](https://discuss.elastic.co/t/crash-regulier/163795 "2019-01-10T18:02:37Z")
**Posts on this page:** 4
**Page:** 1
### Author:  [@cko](https://discuss.elastic.co/u/cko)
#### Post date: [January 10, 2019, 6:02pm UTC](https://discuss.elastic.co/t/crash-regulier/163795/1 "2019-01-10T18:02:37Z")
Hello
Depuis quelque temps, je rencontre des soucis sur mon cluster Elasticsearch que je ne parviens pas à résoudre.
J'ai 5 nodes dans mon cluster :
- 64 Go de ram ( 32 dédiées à la heap)
- CPU 2x20 core
- HDD en 7200 RPM sur 3 des 5 machines
- HDD en 10K RPM pour 2 des 5 machines
J'ai environ 130 indices et 1200 shards.
Chaque indice fait entre 200 Mo et 1 To.
Il arrive très régulièrement (tout les jours à vrai dire) qu'un node plante totalement. Je suis obligé de lancer un reboot, impossible de restart le service.
A la suite de cela, le recovery est très long (jusqu’à 12 heures) ...
Voici le type de log que j'ai régulièrement :
```auto
[2019-01-10T03:38:49,281][INFO][o.e.m.j.JvmGcMonitorService] [elasticsearch-p2] [gc][37497] overhead, spent [501ms] collecting in the last [1.3s]
[2019-01-10T04:45:20,341][INFO][o.e.m.j.JvmGcMonitorService] [elasticsearch-p2] [gc][41385] overhead, spent [425ms] collecting in the last [1.1s]
[2019-01-10T04:54:37,939][INFO][o.e.m.j.JvmGcMonitorService] [elasticsearch-p2] [gc][41930] overhead, spent [419ms] collecting in the last [1.4s]
```
Ou encore :
```auto
019-01-10T17:35:26,074][WARN][o.e.c.s.ClusterService] [elasticsearch-p2] cluster state update task [zen-disco-receive(from master [master {elasticsearch-p3}{fneleMCmS3WT40g9CRVDtw}{sFFOLVQ1Qpq70Urokdeg6w}{10.5.10.13}{10.5.10.13:9300}{rack=elasticsearch-p3} committed version [108378]])] took [58.3s] above the warn threshold of 30s
[2019-01-10T17:38:36,360][WARN][o.e.c.s.ClusterService] [elasticsearch-p2] cluster state update task [zen-disco-receive(from master [master {elasticsearch-p3}{fneleMCmS3WT40g9CRVDtw}{sFFOLVQ1Qpq70Urokdeg6w}{10.5.10.13}{10.5.10.13:9300}{rack=elasticsearch-p3} committed version [108379]])] took [3.1m] above the warn threshold of 30s
```
Avez-vous des pistes ou des recommandations pour orienter mes recherches et résoudre ce problème ?
Merci par avance pour votre aide
Cordialement
---
### Author:  [@dadoonet](https://discuss.elastic.co/u/dadoonet)
#### Post date: [January 10, 2019, 6:22pm UTC](https://discuss.elastic.co/t/crash-regulier/163795/2 "2019-01-10T18:22:59Z")
Tu as sans doute trop de shards par noeud.
Regarde
[https://www.elastic.co/elasticon/conf/2016/sf/quantitative-cluster-sizing](https://www.elastic.co/elasticon/conf/2016/sf/quantitative-cluster-sizing)
> **[How many shards should I have in my Elasticsearch cluster?
| Elastic](https://www.elastic.co/blog/how-many-shards-should-i-have-in-my-elasticsearch-cluster)**
>
> Elasticsearch is a very versatile platform, that supports a variety of use cases, and provides great flexibility around data organisation and replication strategies. This flexibility can however somet...
> **[NetSecureDay: Managing your Black Friday Logs](https://speakerdeck.com/elastic/netsecureday-managing-your-black-friday-logs)**
>
> Surveiller une application complexe n’est pas une tâche aisée, mais avec les bons outils, ce n’est pas si sorcier. Néanmoins, des périodes fortes telles que les opérations de type « Black Friday » (Vendredi noir) ou période de Noël peuvent pousser...
Et [https://www.elastic.co/webinars/using-rally-to-get-your-elasticsearch-cluster-size-right](https://www.elastic.co/webinars/using-rally-to-get-your-elasticsearch-cluster-size-right)
---
### Author:  [@cko](https://discuss.elastic.co/u/cko)
#### Post date: [January 11, 2019, 4:20pm UTC](https://discuss.elastic.co/t/crash-regulier/163795/3 "2019-01-11T16:20:38Z")
Hello,
Merci pour ton retour. Je vais creuser de ce coté 🙂
Cordialement
---
### Author:  [@system](https://discuss.elastic.co/u/system)
#### Post date: [February 8, 2019, 4:20pm UTC](https://discuss.elastic.co/t/crash-regulier/163795/4 "2019-02-08T16:20:42Z")
This topic was automatically closed 28 days after the last reply. New replies are no longer allowed.