Azure应用程序服务计划差异P2V2和P1V3?
我有一个网站在Azure P1V2上运行一段时间,现在它开始在CPU上挣扎(假设它比其他问题更多)。 因此,目前,我正在考虑购买一个更好的计划,在P2V2和P1V3之间非常困惑。
P1V3具有2VCU和更多的内存,但比P2V2便宜,P2V2也具有2个CPU。那么,为什么P2V2更昂贵?
同样,P1V3具有最小的195 ACU,但P2V2总计420个ACU。根据Microsoft Doc,ACU目前在小的(标准_A1)VM上标准化为100,所有其他SKU,然后所有其他SKU表示SKU可以运行标准基准标准的速度大约更快。
由于P1V3的195 ACU最少,但是2 VCPU,这是否意味着我们应该期望最小的390 ACU(195 *2)?那么P2V2会更快但不多?
I have a site running on azure P1v2 for a while and now it starts to struggle on CPU (assume it is more requests than other issues).
So I am looking at buying a better plan, at the moment, quite confused between p2v2 and p1v3.
P1V3 have 2vCU and bit more memory, but it is cheaper than P2V2, which also has 2 CPU. So why p2v2 is more expensive?
Also p1v3 has a minimal 195 ACU but p2v2 has 420 total ACU.According to Microsoft doc, ACU is currently standardized on a Small (Standard_A1) VM being 100 and all other SKUs then represent approximately how much faster that SKU can run a standard benchmark.
As p1v3 has minimal 195 ACU but 2 vCPU, does it mean we should expect minimal 390 ACU (195 *2)? So p2v2 will be faster but not much?
如果你对这篇内容有疑问,欢迎到本站社区发帖提问 参与讨论,获取更多帮助,或者扫码二维码加入 Web 技术交流群。

绑定邮箱获取回复消息
由于您还没有绑定你的真实邮箱,如果其他用户或者作者回复了您的评论,将不能在第一时间通知您!
发布评论
评论(2)
我会这样解释:
该文档说对于V3,此“ ACU每VCPU”,而对于V2,则为“总acu”,
因此在您的情况下:
p2v2:210(per vcpu)x 2 = 420总acu
p1v3:190(每VCPU)x 2 = 380总aCU(最小值 - 见下文),
因为v3包含超线程,而v2则没有,我想实际上190的最小值是将其解释为最糟糕的案例(当而不是当时而不是最差的库2个真正的内核仅使用1个带有2个线程的真实核心)。
我猜想在大多数情况下,V3将具有更多的ACU,并且不能始终保证始终可用(因此是最低ACU)。
I would interpret it as this:
The documentation says that for v3 this "ACU per vCPU", while for v2 it says "total ACU"
So in your case:
P2v2: 210 (per vCPU) x 2 = 420 total ACU
P1v3: 190 (per vCPU) x 2 = 380 total ACU (minimum - see below)
Since v3 contains hyper-threading while v2 does not, I guess the minimal of 190 in fact is to be interpreted as worst-case (when instead of 2 real cores only 1 real core with 2 threads is used).
I would guess that in most cases v3 will have more ACU and that this simply is just not guaranteed to be available at all times (therefore the minimum ACU).
根据 app service service Pricing pricing 说:
p1v3 - 付费 -
$ 0.33/小时
电荷模型
:与V2计划一样,您现在可以使用一个和三年的预订价格, 。 V3计划现在也可以在开发/测试订阅中获得。
p2v2 - 付费 -
$ 0.40/小时
如果您承诺更长的时间,则可以通过预留实例节省更多的钱。
在Linux和Windows上的PV3定价保留实例可让您以比相同的PV2实例运行!
As per the App Service Pricing document, it says:
P1v3 - Pay as you go -
$0.33/hour
Charge model
:As with the v2 Plan, you now have access to one and three-year reservation price, which lowers costs. v3 Plans are now available in dev/test subscriptions as well.
p2v2 - Pay as you go -
$0.40/hour
You will save more money with reserved instances if you commit for a longer period of time.
Reserved instances pricing for Pv3 on Linux and Windows allows you to run at a lower cost than identical Pv2 instances!