BLU Discuss list archive
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[Discuss] khugepaged + vmware = massive CPU load
- Subject: [Discuss] khugepaged + vmware = massive CPU load
- From: richard.pieri at gmail.com (Rich Pieri)
- Date: Mon, 14 Dec 2015 16:56:47 -0500
- In-reply-to: <566F212A.6090508@mattgillen.net>
- References: <22122.13681.176279.74573@snorkack.blazemonger.com> <20151211114214.GH27495@randomstring.org> <20151213155434.b9951c82e986f06526726be0@gmail.com> <566F212A.6090508@mattgillen.net>
On 12/14/2015 3:06 PM, Matthew Gillen wrote: > That explains why hugepages are useful, but it doesn't explain why it > would be detrimental in some cases. The best I could come up with is If you have lots of small allocations then having large pages means either you're going to waste lots of RAM or you're going to run out of pages. It's a lot like putting 50 byte files on a 4GB FAT flash drive with 64K clusters. -- Rich P.
- References:
- [Discuss] khugepaged + vmware = massive CPU load
- From: dbarrett at blazemonger.com (Daniel Barrett)
- [Discuss] khugepaged + vmware = massive CPU load
- From: dsr at randomstring.org (Dan Ritter)
- [Discuss] khugepaged + vmware = massive CPU load
- From: shankar.viswan at gmail.com (Shankar Viswanathan)
- [Discuss] khugepaged + vmware = massive CPU load
- From: me at mattgillen.net (Matthew Gillen)
- [Discuss] khugepaged + vmware = massive CPU load
- Prev by Date: [Discuss] khugepaged + vmware = massive CPU load
- Next by Date: [Discuss] Boston Linux Meeting reminder tomorrow, December 16, 2015 - Introduction to Ceph and Architectural Overview
- Previous by thread: [Discuss] khugepaged + vmware = massive CPU load
- Next by thread: [Discuss] khugepaged + vmware = massive CPU load
- Index(es):