# Pulling projects reloads all flows

**URL:** https://discourse.nodered.org/t/pulling-projects-reloads-all-flows/11170
**Category:** General
**Created:** [15 May 2019 14:31 UTC](https://discourse.nodered.org/t/pulling-projects-reloads-all-flows/11170 "2019-05-15T14:31:13Z")
**Posts on this page:** 3
**Page:** 1

<div class="post-metadata">

### Author: ![joost](https://avatars.discourse-cdn.com/v4/letter/j/6de8d8/32.png) [@joost](https://discourse.nodered.org/u/joost)
#### Post date: [15 May 2019 14:31 UTC](https://discourse.nodered.org/t/pulling-projects-reloads-all-flows/11170/1 "2019-05-15T14:31:13Z")

</div>

Hi all,  
First post, so sorry if I'm missing some essentials here.

I'm wondering if there is an option to **not** redeploy all flows when pulling a "project" from the repo. This is a bit of a risk for the currently running flows because, for example:

- any 'delay' nodes will have their buffers reset (and therefore lose data)
- inject on start will inject again
- variables (global and flow) are reset (although these could be persisted in latest versions I believe?)

Is there an option to do a deploy similar to "only modified nodes" instead? I haven't been able to find this in the documentation. Any pointers in the right direction would also help!

Thanks!

---

<div class="post-metadata">

### Author: ![knolleary](https://sea2.discourse-cdn.com/flex026/user_avatar/discourse.nodered.org/knolleary/32/3_2.png) [@knolleary](https://discourse.nodered.org/u/knolleary)
#### Post date: [15 May 2019 14:34 UTC](https://discourse.nodered.org/t/pulling-projects-reloads-all-flows/11170/2 "2019-05-15T14:34:27Z")

</div>

Hi @joost - no such option exists but we could look at adding one in the future.

---

<div class="post-metadata">

### Author: ![joost](https://avatars.discourse-cdn.com/v4/letter/j/6de8d8/32.png) [@joost](https://discourse.nodered.org/u/joost)
#### Post date: [15 May 2019 14:42 UTC](https://discourse.nodered.org/t/pulling-projects-reloads-all-flows/11170/3 "2019-05-15T14:42:20Z")

</div>

Hi @knolleary! Thanks for the quick response.

Do you know of any suggested options to avoid this, short of writing a delay node that persists data across redeploys instead of in mem?
