解决了

CVM资源

  • 2016年3月29日
  • 9回复
  • 2094意见
b
徽章 +7
我们有一个4主持的戴尔捆绑包。

每个主机：
  • 2 x Xeon CPU E5-2620 V3 @ 2.40GHz
  • 256GB RAM
  • 4.4 TB存储400GB为SSD
运行NCC检查后，我们得到了此错误：

节点10.0.0.58：FAIL：CVM内存16433468 kb小于阈值24 GBNODE 10.0.0.55：FAIL：CVM内存16433468 kb小于阈值24 GBNODE 24 GBNODE 10.0.0.57：FAIL：FAIL：FAIL：CVM MOMEMOR：CVM 1643468 KB BB IS BLESH IS BLESH IS BLESS24 GBNODE 10.0.0.56：FAIL：CVM内存16433468 kb小于阈值24 GBRefer至KB 1513，以获取有关ldap_config_check的CVM_MEMORY_CHECKDETAILED信息的详细信息，请访问NOODE.0.0.56:INFO：no ldap inspeciedy：kb in no ldap inspeciedy：kb。cvm_numa_configuration_check：node 10.0.0.58：info：cvm的ldap_config_checkdetailed信息：cvm上的vcpus（8）数量大于max cores（6）per numa numa node.node.node.node 10.0.0.0.0.55:Info：大于每个numa节点的最大内核（6）。节点10.0.0.57：info：CVM上的VCPU（8）的数量大于每个numa node.node.node.0.0.0.0.56：Inmumb：number：numa numa numa numa numa nume cores（6）CVM上的VCPU（8）的最大核心（6）每个numa节点。引用到KB 3034，以获取有关CVM_NUMA_CONFIGURATION上的详细信息vm_checks的离子：我将所有内容都留为dell的默认设置。我考虑过打电话给戴尔（Dell），但我想我可能更喜欢从马的嘴里听到它。

1.这些是6核处理器，因此CVM是否可以更好地为6VCPU配置？这样他们就不必等待下一个处理器的6个内核和2个来进行指导？

2.该集群的最佳内存设置可能是什么？这是一些设置：

  • 复制因子2
  • 压缩
  • 压缩延迟0分钟。
  • 压缩空间保存了1.3 TIB
  • 压缩比1.53：1
  • 性能层删除
  • 磁盘重复数据删除
  • 在磁盘上保存空间上 -
  • 在磁盘defup比率1：1
图标

最好的答案乔恩2016年3月31日，11：03

\n
\nI'll add a bit of color here
\n
\nRE Memory
\nNCC is flagging the memory because you've got perf-tier Dedupe on, which is recommended to have 24GB of CVM memory per host for.
\n
\nInline compression is fantastic and only requires 16GB memory, so no worries there.
\n
\nRE CPU's
\nSuper, super, super long story, the elevator version is that for the longest time, our \"default\" has been 8 vCPU CVM's, but we dont RESERVE all 8 vCPU's, so the hypervisor scheduler will prioritize along with \"user\" VM's (servers, etc).
\n
\nThat NCC check was actually partially driven by me, and to be frank, it went out the door sooner than we wanted. a future NCC release (likely 2.2.2, coming out shortly) will disable this check to prevent confusion in the field.
\n
\nThat said, going forward, a future, future NCC\/NOS release will do the following
\nTurn that check back on
\nGive proper guidance on CVM sizing based on the hardware platform
\nDefaults for \"new\" imaging will take these guidances into account, so it will be good from the jump on new systems. Existing systems would need a bit of tweaking of course.
\n
\n
\n
\nBottomline: Either turn off perf tier dedupe to get rid of the first warning (or increase CVM to 24G), and ignore that CVM NUMA check for now, until the new guidance comes out","className":"post__content__best_answer"}">
查看原件
    \n
    \nPer Host:
    \n
      \n
    • 2 x Xeon CPU E5-2620 v3 @ 2.40GHz \n
    • 256GB RAM \n
    • 4.4 TB of storage 400GB of which is SSD<\/ul>we got this error after running the NCC checks:
      \n
      \nNode 10.0.0.58:FAIL: CVM memory 16433468 kB is less than the threshold 24 GBNode 10.0.0.55:FAIL: CVM memory 16433468 kB is less than the threshold 24 GBNode 10.0.0.57:FAIL: CVM memory 16433468 kB is less than the threshold 24 GBNode 10.0.0.56:FAIL: CVM memory 16433468 kB is less than the threshold 24 GBRefer to KB 1513 for details on cvm_memory_checkDetailed information for ldap_config_check:Node 10.0.0.56:INFO: No ldap config is specified.Refer to KB 2997 for details on ldap_config_checkDetailed information for cvm_numa_configuration_check:Node 10.0.0.58:INFO: Number of vCPUs(8)on the CVM is more than the max cores(6) per NUMA node.Node 10.0.0.55:INFO: Number of vCPUs(8)on the CVM is more than the max cores(6) per NUMA node.Node 10.0.0.57:INFO: Number of vCPUs(8)on the CVM is more than the max cores(6) per NUMA node.Node 10.0.0.56:INFO: Number of vCPUs(8)on the CVM is more than the max cores(6) per NUMA node.Refer to KB 3034 for details on cvm_numa_configuration_checkDetailed information for vm_checks:I left everything as default from Dell. I thought about calling Dell about this config, but I thought I might prefer to hear it from the horse's mouth.
      \n
      \n1. these are 6-core processors, so wouldn't the CVM's be better configured for 6vCPU's? so that they don't have to wait on 6 cores and 2 from the next processor to do instruction?
      \n
      \n2. What might be the optimum memory settings for this Cluster? here's some settings:
      \n
      \n
        \n
      • Replication Factor 2 \n
      • Compression On \n
      • Compression Delay 0 Min. \n
      • Compression Space Saved 1.3 TiB \n
      • Compression Ratio 1.53 : 1 \n
      • Performance Tier Deduplication On \n
      • On Disk Deduplication Off \n
      • On Disk Dedup Space Saved - \n
      • On Disk Dedup Ratio 1 : 1<\/ul>","quoteUsername":"bpr-admin","translations":{"Common":{"like":"Like","unlike":"Unlike"},"Forum":{"Quote":"Quote","Share":"Share"}}}">
    该主题已关闭以供评论

    9回复

    b
    UserLevel 4
    徽章 +17
    默认CVM配置使用8 VCPU和16GB内存

    NCC将在每个Nutanix最佳实践中运行，以告知群集状况
    随着压缩和dedupe的打开，最好将24GB分配给CVM
    \n
    \nNCC will run per Nutanix best practice to inform cluster condition
    \nAs Compress and Dedupe is on, better assign 24GB to the CVM","quoteUsername":"bezeddin","translations":{"Common":{"like":"Like","unlike":"Unlike"},"Forum":{"Quote":"Quote","Share":"Share"}}}">
    b
    徽章 +7
    我很好奇Nutanix可能会为Hexa核处理器进行优化。我知道默认值为8 VCPU。问题是关于系统优化。有什么见识吗？
    b
    UserLevel 4
    徽章 +17
    虚拟处理器调度算法将优化CVM上的CPU利用率
    CVM就像VM在虚拟机构上运行
    CPU核心始终动态更改
    \nCVM just like VM running on Hypervisor
    \nCPU core utilization dynamically change all the time","quoteUsername":"bezeddin","translations":{"Common":{"like":"Like","unlike":"Unlike"},"Forum":{"Quote":"Quote","Share":"Share"}}}">
    乔恩
    秩
    UserLevel 6
    徽章 +29
    嗨，谢谢您的伸出手。

    我在这里添加一点颜色

    RE内存
    NCC正在标记内存，因为您已经启用了perf-ter dedupe，建议每个主机的CVM内存24GB。

    内联压缩很棒，只需要16GB的内存，因此不用担心。

    Re CPU
    超级，超级长故事，电梯版本是最长的时间，我们的“默认值”是8个VCPU CVM，但我们不保留所有8个VCPU，因此Hypervisor Scheduler将与“用户” VM一起优先考虑（服务器）（服务器， ETC）。

    NCC检查实际上是由我部分驱动的，坦率地说，它比我们想要的要早。未来的NCC版本（可能很快出现2.2.2）将禁用此支票，以防止现场混淆。

    也就是说，展望未来，未来的NCC/NOS发布将执行以下操作
    重新打开检查
    根据硬件平台提供有关CVM尺寸的适当指导
    “新”成像的默认值将考虑这些指南，因此从新系统的跳跃中可以很好。现有系统当然需要进行一些调整。



    底线：要么关闭perf tier Dedupe，以摆脱第一个警告（或将CVM提高到24克），然后忽略CVM Numa检查，直到新指南出现
    乔恩·科勒|Nutanix工程技术总监|Nutanix NPX＃003，VCDX＃116 |@jonkohler |如果有用的话，请荣誉！
    \n
    \nI'll add a bit of color here
    \n
    \nRE Memory
    \nNCC is flagging the memory because you've got perf-tier Dedupe on, which is recommended to have 24GB of CVM memory per host for.
    \n
    \nInline compression is fantastic and only requires 16GB memory, so no worries there.
    \n
    \nRE CPU's
    \nSuper, super, super long story, the elevator version is that for the longest time, our \"default\" has been 8 vCPU CVM's, but we dont RESERVE all 8 vCPU's, so the hypervisor scheduler will prioritize along with \"user\" VM's (servers, etc).
    \n
    \nThat NCC check was actually partially driven by me, and to be frank, it went out the door sooner than we wanted. a future NCC release (likely 2.2.2, coming out shortly) will disable this check to prevent confusion in the field.
    \n
    \nThat said, going forward, a future, future NCC\/NOS release will do the following
    \nTurn that check back on
    \nGive proper guidance on CVM sizing based on the hardware platform
    \nDefaults for \"new\" imaging will take these guidances into account, so it will be good from the jump on new systems. Existing systems would need a bit of tweaking of course.
    \n
    \n
    \n
    \nBottomline: Either turn off perf tier dedupe to get rid of the first warning (or increase CVM to 24G), and ignore that CVM NUMA check for now, until the new guidance comes out","quoteUsername":"Jon","translations":{"Common":{"like":"Like","unlike":"Unlike"},"Forum":{"Quote":"Quote","Share":"Share"}}}">
    b
    徽章 +7
    我有很多记忆，所以我将其带到24 GB。

    但是，在看到NUMA检查之前，我确实认为让CVM在处理器为6核的系统上需要8VCPU是很奇怪的。我对VMware的理解是，为了使指令传递给CPU，需要8个内核，并且直到它们为VM必须等待。我对这种理解不正确吗？
    \n
    \nHowever, before ever seeing the NUMA check, I did think it odd to have the CVM require 8vCPU on a system where the processors are 6-core. My understanind of VMware is that in order for an instruction to pass to the CPU, 8 cores would need to be available, and that until they are the VM has to wait. Am I incorrect in this understanding?","quoteUsername":"bpr-admin","translations":{"Common":{"like":"Like","unlike":"Unlike"},"Forum":{"Quote":"Quote","Share":"Share"}}}">
    乔恩
    秩
    UserLevel 6
    徽章 +29
    您正在考虑共同开展和共同营业，这实际上并不比ESX 3.x天的5.5和6.0中的问题大。


    这里的问题是“ numa width”而不是共同停留，这不是CPU调度问题，而更多的是内存调度问题，从理论上讲，您可以使CPU线程在处理器之间的QPI上解决QPI，该问题是在处理器之间的内存，不是死刑，只是不理想。

    所有这些Numa的东西都是尽可能优化的。您现在可以使用配置了，在将来发布后，您可能会看到指导的变化，在我们完成一些工程和质量检查方面的努力使足迹较小后，这应该解决6个核心系统上的NUMA宽度。
    乔恩·科勒|Nutanix工程技术总监|Nutanix NPX＃003，VCDX＃116 |@jonkohler |如果有用的话，请荣誉！
    \n
    \n
    \nthe issue here is \"NUMA width\" rather than co-stop, which is less of an CPU scheduling issue, and more of a memory scheduling issue, where you could, in theory, have CPU threads addressing memory across the QPI between processors, which isn't a death sentence, its just less than ideal.
    \n
    \nAll of this NUMA stuff is all about optimizing as much as possible. You'll be fine with your config now,<\/b> and you'll likely see a change in guidance after a future release, after we finalize some engineering and QA efforts on making the footprint smaller, which should address NUMA width on 6 core systems.","quoteUsername":"Jon","translations":{"Common":{"like":"Like","unlike":"Unlike"},"Forum":{"Quote":"Quote","Share":"Share"}}}">
    b
    徽章 +7
    谢谢，那是非常有用的信息！
    b
    徽章 +7
    关于此的另一个问题：

    我正要开始将24GB应用于CVM的内存大小，但是，当我查看活动RAM时，CVM都在4或5 GB范围内。24GB真的是最好的吗？
    \n
    \nI was about to start applying 24GB to the memory sizes of the CVM however, when I looked at Active RAM the CVM's are all either in the 4 or 5 GB range. Is 24GB really the best?","quoteUsername":"bpr-admin","translations":{"Common":{"like":"Like","unlike":"Unlike"},"Forum":{"Quote":"Quote","Share":"Share"}}}">
    乔恩
    秩
    UserLevel 6
    徽章 +29
    主动内存将取决于您的努力驱动工作负载。现在可能不是很高，但是在更高利用率的时候，随着您增加更多的工作量和需求，它将上升。

    过去，请记住，我们将尽可能保持该记忆，以通过将尽可能多的数据保留在其中。
    乔恩·科勒|Nutanix工程技术总监|Nutanix NPX＃003，VCDX＃116 |@jonkohler |如果有用的话，请荣誉！
    \n
    \nPast that, Keep in mind that we'll keep that memory as full as possible, to promote cache hits by keeping as much data in there as possible.","quoteUsername":"Jon","translations":{"Common":{"like":"Like","unlike":"Unlike"},"Forum":{"Quote":"Quote","Share":"Share"}}}">
    条款和条件
    Learn more about our cookies.<\/a>","cookiepolicy.button":"Accept cookies","cookiepolicy.button.deny":"Deny all","cookiepolicy.link":"Cookie settings","cookiepolicy.modal.title":"Cookie settings","cookiepolicy.modal.content":"We use 3 different kinds of cookies. You can choose which cookies you want to accept. We need basic cookies to make this site work, therefore these are the minimum you can select. Learn more about our cookies.<\/a>","cookiepolicy.modal.level1":"Basic
    Functional","cookiepolicy.modal.level2":"Normal
    Functional + analytics","cookiepolicy.modal.level3":"Complete
    Functional + analytics + social media + embedded videos"}}}">
    Baidu