# For improving IPFS GET performance

**URL:** <https://discuss.ipfs.tech/t/for-improving-ipfs-get-performance/4119>\
**Category:** Help\
**Created:** [October 23, 2018, 6:56am UTC](https://discuss.ipfs.tech/t/for-improving-ipfs-get-performance/4119 "2018-10-23T06:56:56Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![endeavor](https://avatars.discourse-cdn.com/v4/letter/e/7ba0ec/32.png) [@endeavor](https://discuss.ipfs.tech/u/endeavor)\
**Post date:** [October 23, 2018, 6:56am UTC](https://discuss.ipfs.tech/t/for-improving-ipfs-get-performance/4119/1 "2018-10-23T06:56:56Z")

</div>

Hi, everyone.  
I’m studying IPFS.  
I found something to share, but it may be my misunderstanding IPFS.

In the source, responseemitter.go([go-ipfs-cmds/http/responseemitter.go at f4d1d56f3d96c917f546de507422dc2258d89a62 · ipfs/go-ipfs-cmds · GitHub](https://github.com/ipfs/go-ipfs-cmds/blob/f4d1d56f3d96c917f546de507422dc2258d89a62/http/responseemitter.go#L280)) and extractor.go ([tar-utils/extractor.go at 8c6c8ba81d5c71fd69c0f48dbde4b2fb422b6dfc · whyrusleeping/tar-utils · GitHub](https://github.com/whyrusleeping/tar-utils/blob/8c6c8ba81d5c71fd69c0f48dbde4b2fb422b6dfc/extractor.go#L175)), there are hard-coded value, 4096.

After changing this value to 256k(in fact, 262144), I tested and found the IPFS GET performance improvement (upto 720MB/s) in small- sized-file(5.5MB) case, but not big-sized-file case.

I guess this value can effect the throughput of download in various context, and must be a member in the configuration.