TE
TechEcho
Home24h TopNewestBestAskShowJobs
GitHubTwitter
Home

TechEcho

A tech news platform built with Next.js, providing global tech news and discussions.

GitHubTwitter

Home

HomeNewestBestAskShowJobs

Resources

HackerNews APIOriginal HackerNewsNext.js

© 2025 TechEcho. All rights reserved.

Meltdown Initial Performance Regressions

226 pointsby natbobcover 7 years ago

9 comments

mrepover 7 years ago
&gt; Applications that have high syscall rates include proxies, databases, and others that do lots of tiny I&#x2F;O. Also microbenchmarks, which often stress-test the system, will suffer the largest losses.<p>My team&#x27;s RDS instances got hit hard with a 40% increase in CPU usage: <a href="https:&#x2F;&#x2F;imgur.com&#x2F;a&#x2F;khGxU" rel="nofollow">https:&#x2F;&#x2F;imgur.com&#x2F;a&#x2F;khGxU</a>
评论 #16346537 未加载
评论 #16347174 未加载
scott_sover 7 years ago
The syscall in his benchmark made me laugh (<a href="https:&#x2F;&#x2F;github.com&#x2F;brendangregg&#x2F;Misc&#x2F;blob&#x2F;master&#x2F;s1bench&#x2F;s1bench.c#L124" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;brendangregg&#x2F;Misc&#x2F;blob&#x2F;master&#x2F;s1bench&#x2F;s1b...</a>):<p><pre><code> close(999); &#x2F;&#x2F; the syscall (it errors, but so what)</code></pre>
评论 #16344985 未加载
jnordwickover 7 years ago
An open question is still who should enable the mitigation. The risk-cost doesn&#x27;t seem to fit too many scenarios.<p>Meltdown requires running native, intrusted code, and that doesn&#x27;t apply to too many servers. While it may be possible to chain this onto another exploit, once an attacker has gained remote code execution, you have much bigger problems.<p>While meltdown is interesting, i wouldn&#x27;t enable kpti on my database servers buried behind other network infrastructure.
评论 #16346288 未加载
Scaevolusover 7 years ago
What syscall rates do different databases sustain at maximum load? Transparent huge pages negating most of the overhead is very good news-- but probably helps less with mmap&#x27;d IO which so many databases use.
评论 #16345282 未加载
vbernatover 7 years ago
It would be interesting to know the interaction between the patched host and the patched guest. As a simple example, if the host aggressively flushes the TLB, the performance impact on the guest of doing the same could be lower. On the other hand, depending on how the host was patched, the loss of performance of the guest could be different when using some features.
rmrfrmrfover 7 years ago
Out of curiosity, why does the syscall rate scale descend from left to right?
评论 #16346911 未加载
voidlogicover 7 years ago
It would be great to see performance deltas for AMD CPUs too, especially since Meltdown only effects Intel and AMD patches for Spectre Variant 2 are considered optional. I would also be nice to see a discussion of AMD&#x27;s ASID and any differences it has with Intel&#x27;s PCID when when PCID is addressed.
评论 #16345466 未加载
b4lancesh33tover 7 years ago
I&#x27;m kinda surprised Netflix weren&#x27;t already using THP.
评论 #16346541 未加载
评论 #16346502 未加载
评论 #16347130 未加载
kev009over 7 years ago
Do you have any idea how much it will regress your I&#x2F;O workload on OCA?
评论 #16346463 未加载