# mediasoup worker memory do not released

**URL:** <https://mediasoup.discourse.group/t/mediasoup-worker-memory-do-not-released/5527>\
**Category:** Deployment & Scalability\
**Created:** [September 15, 2023, 8:42am UTC](https://mediasoup.discourse.group/t/mediasoup-worker-memory-do-not-released/5527 "2023-09-15T08:42:33Z")\
**Posts on this page:** 4\
**Page:** 1

<div class="post-metadata">

**Author:** ![gonodefunc](https://yyz2.discourse-cdn.com/free1/user_avatar/mediasoup.discourse.group/gonodefunc/32/2345_2.png) [@gonodefunc](https://mediasoup.discourse.group/u/gonodefunc)\
**Post date:** [September 15, 2023, 8:42am UTC](https://mediasoup.discourse.group/t/mediasoup-worker-memory-do-not-released/5527/1 "2023-09-15T08:42:33Z")

</div>

When I was doing stress testing with mediasoup, I found that the memory usage of mediasoup-worker kept increasing when there were many consumers and the network bandwidth for consuming video reached the limit. After closing the room and all routers, the memory used by mediasoup-worker was not released. I cannot provide memory-related logs because if I start VALGRIND, the CPU usage will be very high and I cannot perform stress testing.

I would like to know why mediasoup-worker occupies so much memory and why the memory is not released after closing all routers?

I am experiencing the same issue when using a single router.

thx all.

mediasoup version 3.11.26

 ![mul-router](https://global.discourse-cdn.com/free1/uploads/mediasoup/original/2X/1/1305bde4ed263072843d0fd718214af05dea70ac.jpeg)

---

<div class="post-metadata">

**Author:** ![BronzedBroth](https://avatars.discourse-cdn.com/v4/letter/b/ad7895/32.png) [@BronzedBroth](https://mediasoup.discourse.group/u/BronzedBroth)\
**Post date:** [September 15, 2023, 9:51pm UTC](https://mediasoup.discourse.group/t/mediasoup-worker-memory-do-not-released/5527/2 "2023-09-15T21:51:16Z")

</div>

Doesn’t sound normal at all, I’m using a much older version currently; does going back years/months do anything?

---

<div class="post-metadata">

**Author:** ![gonodefunc](https://yyz2.discourse-cdn.com/free1/user_avatar/mediasoup.discourse.group/gonodefunc/32/2345_2.png) [@gonodefunc](https://mediasoup.discourse.group/u/gonodefunc)\
**Post date:** [September 18, 2023, 1:55am UTC](https://mediasoup.discourse.group/t/mediasoup-worker-memory-do-not-released/5527/3 "2023-09-18T01:55:32Z")

</div>

Did your network bandwidth also reach its limit?

---

<div class="post-metadata">

**Author:** ![vuzya](https://yyz2.discourse-cdn.com/free1/user_avatar/mediasoup.discourse.group/vuzya/32/2472_2.png) [@vuzya](https://mediasoup.discourse.group/u/vuzya)\
**Post date:** [December 19, 2023, 1:23pm UTC](https://mediasoup.discourse.group/t/mediasoup-worker-memory-do-not-released/5527/4 "2023-12-19T13:23:27Z")

</div>

Same issue here. Can reproduce even for single viewer - the key point is viewer (or server itself) should have not enough bandwidth. It looks like a backpressure issue. @gonodefunc have you fixed it somehow?