In this episode, Sean Donahue is joined by Al Solorzano of E360 to discuss the worst disasters they've recovered from. Surprise twist, not all clouds are cumulonimbus and not all disasters are natural.<\/p><\/oembed>

Resources<\/p>

最大SSD与HDD比率

  • 2016年8月23日
  • 4个答复
  • 1630次观看
一种
徽章 +7
只是想知道以下内容：您所处的任何地方都会显示2或4 SSD的Nutanix配置，其余为HDD。随着在每个型号上都有所有闪存节点的进步，是否有任何阻止我们说6 SSD和4 HDD？还是每个支持10个磁盘的节点中的5个？Pro许可也使我们能够将VM或部分固定在SSD上？
    该主题已关闭以供评论

    4个答复

    n
    Userlevel 1
    徽章 +4
    这只是我的.02，但总的来说，混合动力车背后的想法是，您的活跃数据集可以适合SSD层，并有一点点的头部空间，以便最具成本效益，而我想请注意，有时您会注意到您可能在或不移动到SSD之前/没有将其击中一些很少使用的数据。尽管有了所有的闪光灯，您要么需要/需要保证IO就会出现在SSD上，而且要花很多钱可以花。我的猜测是，如果您开始走向50/50 SSD/HDD或60/40 SSD/HDD比率，您可能无论如何都可能接近All Flash配置的价格，所以我想您只会得到一点多亏了HDD层。



    在磁盘性能成为问题之前，您会用完CPU和内存吗？我经常做。如果是这样，那么也许需要更多的节点来从CPU和内存的角度来支持工作负载，并且您可能真的不需要HDD层中的额外容量，或者您可能要定位的Flash量。



    有趣的想法。

    \n

    \nWill you run out of CPU and memory before disk performance becomes an issue? I always do. If so then perhaps more nodes will be required anyway to support the workload from a CPU & memory perspective and you might not truly need the extra capacity in the HDD tier, or maybe the amount of flash you're targeting.

    \n

    \nInteresting idea though.","quoteUsername":"ndolson","translations":{"Common":{"like":"Like","unlike":"Unlike"},"Forum":{"Quote":"Quote","Share":"Share"}}}">
    一种
    徽章 +7
    SSD功能的PIN呢？
    一种
    徽章 +7
    现在，所有模型都支持所有Flash，是否还会有PIN到SSD的PRO许可证，还是仍然限于最终？



    另外，如果您必须在这两个配置之间进行选择，您会选择什么？



    3 x节点：

    6 x 400GB SSD

    4 x 2TB SATA



    要么



    3 x节点：

    4 x 400GB SSD

    6 x 2TB SATA

    \n

    \nAlso if you had to choose between this two config what would you choose?

    \n

    \n3 x Nodes:

    \n6 x 400GB SSD

    \n4 x 2TB SATA

    \n

    \nor

    \n

    \n3 x Nodes:

    \n4 x 400GB SSD

    \n6 x 2TB SATA","quoteUsername":"alexxdavid","translations":{"Common":{"like":"Like","unlike":"Unlike"},"Forum":{"Quote":"Quote","Share":"Share"}}}">
    乔恩
    秩
    UserLevel 6
    徽章 +29
    要向我们的产品管理团队发送笔记进行三重检查，但我几乎可以肯定的是，“ VM Flash模式”将保持最终功能。



    恕我直言，虽然它对于诸如服务提供商用例之类的东西都是整洁的，但我认为这不适用于企业，尤其是因为现在所有的闪光灯都如此普遍。我宁愿让您在SSD主轴计数上更高，而不是HDD主轴计数，如果您关注群集中的“热数据”量，而不是尝试使用VM闪存模式将现有闪光灯分配（即PIN）到SSD）。



    当然，这是假设您没有能力违背能力，或者您的容量通过降低数据诸如在线压缩（我们强烈建议几乎所有工作负载）进行护理，并且它也可以在SSD和HDD上使用）。



    也就是说，很好奇，这个针对SSD问题是否有特定的驱动程序（即绩效问题，企业需求，特定应用程序的需求等）？



    乔恩
    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

    \nIMHO, while it is neat for things like service provider use cases, I dont see this widely applicable to the enterprise, especially since All Flash is so pervasive now. I'd rather have you go higher on SSD spindle count, rather than HDD spindle count if you are concerned with the amount of \"hot data\" within a cluster, rather than you trying to partition the existing flash with VM Flash Mode (i.e Pin to SSD).

    \n

    \nThis of course assumes that you are not capacity contrained, or your capacity is otherwise taken care of through data reduction techniques like in-line compression (which we highly recommend on just about any workload, and it works across SSD and HDD too).

    \n

    \nThat said, just curious, is there a specific driver for this Pin to SSD question (i.e performance issue, demand from the business, demand from a specific application, etc)?

    \n

    \nJon","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