# Memory usage in Zammad

**URL:** https://community.zammad.org/t/memory-usage-in-zammad/373
**Category:** Technical assistance
**Created:** [March 5, 2018, 4:38pm UTC](https://community.zammad.org/t/memory-usage-in-zammad/373 "2018-03-05T16:38:33Z")
**Posts on this page:** 1
**Showing post:** 7

<div class="post-metadata">

### Author: ![anon29869905](https://community.zammad.org/letter_avatar_proxy/v4/letter/a/b782af/32.png) [@anon29869905](https://community.zammad.org/u/anon29869905)
#### Post date: [March 28, 2018, 7:32am UTC](https://community.zammad.org/t/memory-usage-in-zammad/373/7 "2018-03-28T07:32:17Z")

</div>

Hi @slavian87 - how many concurrent users (agents and customers) work in your Zammad simultaneously at peak time?

Our recommendation for 10 concurrent users and a smooth user experience is:

- 6 CPUs
- 12 GB RAM (because of the hungry ES)
- Put `/opt/zammad/tmp` into a RAM-Disk (2-3 GB, keep an eye on it and extend if required)

For the application server:  
[https://docs.zammad.org/en/latest/appendix-configure-env-vars.html](https://docs.zammad.org/en/latest/appendix-configure-env-vars.html) -\>

`zammad config:set WEB_CONCURRENCY=3`  
`systemctl restart zammad`

One scheduler instance is sufficient

To reduce the memory load of the Ruby processes you can set the Ruby specific ENV vars (as described above):

`export RUBY_GC_MALLOC_LIMIT=1077216 RUBY_GC_MALLOC_LIMIT_MAX=2177216 RUBY_GC_OLDMALLOC_LIMIT=3000100 RUBY_GC_OLDMALLOC_LIMIT_MAX=4000100`

---

_[View the full topic](https://community.zammad.org/t/memory-usage-in-zammad/373)._