blocks|key|1134606|text|您会想要使用大而慢的RAM。与RAM性能和磁盘性能之间的差异相比，RAM性能的差异可以忽略不计。|type|unstyled|depth|inlineStyleRanges|entityRanges|data|entityMap^0^^$0|@$1|2|3|4|5|6|7|C|8|@]|9|@]|A|$]]]|B|$]]

You will want to go with the large and slow RAM. The difference in RAM performance is negligible compared to the difference between RAM performance and disk performance.

blocks|key|3443296|text|好吧，很简单：|type|unstyled|depth|inlineStyleRanges|entityRanges|data|3443297|您的数据库是否适合在48+in的RAM与操作系统和所有？如果是的话拿着吧。否则，就拿96|3443298|此外，数据库适合在中，意味着它符合索引，视图和所有这些。|3443299|SSD的评论完全是胡说八道，带宽和访问时间都不一样，任何SSD都不能证明占用较少的RAM是合理的。|entityMap^0|0|0|0^^$0|@$1|2|3|4|5|6|7|I|8|@]|9|@]|A|$]]|$1|B|3|C|5|6|7|J|8|@]|9|@]|A|$]]|$1|D|3|E|5|6|7|K|8|@]|9|@]|A|$]]|$1|F|3|G|5|6|7|L|8|@]|9|@]|A|$]]]|H|$]]

Alright, it's very very very simple :

Does your database fit in 48GB of RAM with OS and all ? if yes, take that.
Else, take 96GB

Also, database fitting in xyz GB of RAM means it fits with index's, views and all that.

SSD comments are complete utter nonsense, both the bandwidth and access time are not on the same level and no SSD can justify taking less RAM.

blocks|key|3443076|text|只有数据库？根据数据库的不同，我认为较大的RAM会更好。速度差异已被证明是微小的，在最好的，但额外的48+to将/可能产生巨大的变化。|type|unstyled|depth|inlineStyleRanges|entityRanges|data|entityMap^0^^$0|@$1|2|3|4|5|6|7|C|8|@]|9|@]|A|$]]]|B|$]]

Database only? Depending on the database, I would think the larger RAM would be better. Speed difference has been proven to be miniscule at best, but the additional 48gb will / may make a huge difference.

blocks|key|1134705|text|绝对是大内存速度是该死的。|type|unstyled|depth|inlineStyleRanges|entityRanges|data|1134706|从XX世纪90年代开始，随机存取RAM技术的数据在100+ns以下。这使用的是几乎古老的芯片，甚至不适合任何当代的边缘。|1134707|访问随机数据的尖端15k转速硬盘驱动器是以毫秒为单位的。100+ns比1ms短10+000倍(纳米->微->米)。当前RAM更快，HDD需要几毫秒才能访问数据。我不在乎我的内存是否比硬盘快50+000倍，或者比硬盘快30+000倍，如果我能得到更多。|entityMap^0|0|0^^$0|@$1|2|3|4|5|6|7|G|8|@]|9|@]|A|$]]|$1|B|3|C|5|6|7|H|8|@]|9|@]|A|$]]|$1|D|3|E|5|6|7|I|8|@]|9|@]|A|$]]]|F|$]]

Definitely large RAM, speed be damned. 

Access to random data for RAM technology from XX century '90 is below 100 ns. That's using practically ancient chips that won't even physically fit into anything borderline contemporary.

Access to random data for cutting edge 15k rpm hard drives is in measured in miliseconds. 100 ns is 10 000 times shorter (nano -> micro -> milli) than 1 ms. Current RAM is faster, and HDD needs several milliseconds to access data. I couldn't care less if my RAM was 50 000 faster or only 30 000 times faster than HDD, if I could get more.

blocks|key|3443228|text|你必须注意以下几点：|type|unstyled|depth|inlineStyleRanges|entityRanges|data|3443229|内存内存速度取决于两个因素:总线速度和延迟。通常情况下，密度越大的芯片的延迟时间越长，最终意味着速度越慢。|unordered-list-item|3443230|总索引数据是将整个索引数据加载到内存中的最关键的y。索引数据是您在内存中需要的最关键的数据(性能方面的更高的惩罚效应)。|3443231|磁盘速度，你有数据库数据存储在SSD中吗？如果答案是肯定的，特别要注意记忆延迟。|entityMap^0|0|0|0^^$0|@$1|2|3|4|5|6|7|J|8|@]|9|@]|A|$]]|$1|B|3|C|5|D|7|K|8|@]|9|@]|A|$]]|$1|E|3|F|5|D|7|L|8|@]|9|@]|A|$]]|$1|G|3|H|5|D|7|M|8|@]|9|@]|A|$]]]|I|$]]

You must take your attention in some points:

<ul>
<li>Memory lantecy Memory speed depends on two factors: bus speed and latency. Usally chips with more density result in a higher latency, wich finaly means less speed</li>
<li>Total index data The most critical y to load the entire index data into memory. Index data is the most critical data you need in memory (higher penalty effect in performance).</li>
<li>Disk speed Do you have the DB data stored in SSD? If the answer is yes, take care specially of memory latency.</li>
</ul>

blocks|key|3443163|text|内存带宽=/=速度！|type|unstyled|depth|inlineStyleRanges|entityRanges|data|3443164|最重要的缺失信息可能是内存时间和CPU/FSB类型。将CPU内存负载延迟降低几个周期，您将在某些计算中将ram的带宽提高一倍。一些数据库由于操作系统和技术原因而不使用大量的ram，您使用的是哪个数据库服务器？CPU类型？L123缓存级别？要运行的查询类型？数据库的大小？|entityMap^0|0^^$0|@$1|2|3|4|5|6|7|E|8|@]|9|@]|A|$]]|$1|B|3|C|5|6|7|F|8|@]|9|@]|A|$]]]|D|$]]

MEMORY BANDWIDTH =/= SPEED!

Probably the most important piece of missing information information is the memory timings and CPU/FSB type. lower the CPU memory load delay by a few cycles and you will blast ram double the bandwidth in certain calculations. Some databases dont use massive amounts of ram due to operating system and technical reasons, what database server are you using? CPU type? L[123] cache levels? type of queries to be run? size of the database?

blocks|key|1134816|text|在花太多钱买错误的硬件之前，我会在购买硬件之前做一些测试和分析。|type|unstyled|depth|inlineStyleRanges|entityRanges|data|1134817|首先，想想你的SLA。|unordered-list-item|1134818|对性能和响应时间有任何严格的要求吗？|1134819|您的选择应该取决于许多因素：|1134820|在不同的工作负载和使用情况下，瓶颈究竟是什么？|1134821|CPU，内存，存储，网络？|1134822|在更快的存储上花更多的钱比用更多的内存更重要吗？|1134823|CPU比内存更快？更快的网络？软件/sql的轻微重新设计？|1134824|您的分析也可能与开发人员、数据库和软件架构以及sql查询设计人员密切相关。|1134825|如果您正在使用windows+-您可以很容易地运行perfmon来查看当前运行系统的一些静力学，而市长非常幸运地了解了您的需求。|entityMap^0|0|0|0|0|0|0|0|0|0^^$0|@$1|2|3|4|5|6|7|V|8|@]|9|@]|A|$]]|$1|B|3|C|5|D|7|W|8|@]|9|@]|A|$]]|$1|E|3|F|5|D|7|X|8|@]|9|@]|A|$]]|$1|G|3|H|5|6|7|Y|8|@]|9|@]|A|$]]|$1|I|3|J|5|D|7|Z|8|@]|9|@]|A|$]]|$1|K|3|L|5|D|7|10|8|@]|9|@]|A|$]]|$1|M|3|N|5|D|7|11|8|@]|9|@]|A|$]]|$1|O|3|P|5|D|7|12|8|@]|9|@]|A|$]]|$1|Q|3|R|5|D|7|13|8|@]|9|@]|A|$]]|$1|S|3|T|5|6|7|14|8|@]|9|@]|A|$]]]|U|$]]

Before spending too much money on wrong hardware, I would do some tests and analysis before buying hardware.

<ul>
<li>First of all, think about your SLA.</li>
<li>any hard requirements to performance and response times?</li>
</ul>

You choice should depend on many factors:

<ul>
<li>under different workloads and usages, what is actually the bottleneck?</li>
<li>CPU, memory, storage, network?</li>
<li>Is it perhaps more important to spend more money on faster storage than more memory?</li>
<li>faster CPU than more memory? faster network? minor redesign on sofware/sql?</li>
<li>you analysis could also be very much relevant to the developers, database and software architcts, and sql query designers.....</li>
</ul>

If you are using windows - you could easily run perfmon to see some statics on the current running system and mayby be very lucky to get a clear idea of your your needs.

We are currently designing our new database servers, and have come up with a trade off I'm not entirely sure of how to answer.

These are our options: 48GB 1333MHz, or 96GB 1066MHz.

My thinking is that RAM should be plentiful for a Database Server (we have plenty and plenty of data, and some very large queries) rather than as quick as it could be. Apparently we can't get 16GB chips at 1333MHz, hence the choices above.

So, should we get lots of slower RAM, or less faster RAM?

<h2>Extra Info:</h2>

Number of DIMM Slots Available: 6 
Servers: Dell Blades
CPU: 6 core (only single socket due to Oracle licensing).

Database server: Small quick RAM or large slow RAM?

翻译质量差，导致语言生硬或混乱。

没有提供实际的解决方法或示例。

解答不清晰，无法理解或解决问题。

页面排版不美观，阅读体验差。

文章

问答

视频

学习中心

腾讯云实验室

直播

竞赛

腾讯云代码分析专区

腾讯iOA零信任安全管理系统专区

腾讯云架构师技术同盟交流圈

腾讯云数据库专区

腾讯云顾问专区

腾讯云原生专区

腾讯混元专区

腾讯云TCE专区

腾讯云Lighthouse专区

腾讯云HAI专区

腾讯云Edgeone专区

腾讯云存储专区

腾讯云智能专区

腾讯轻联专区 

腾讯云开发专区

TAPD专区

腾讯轻量云游戏服专区

腾讯云最具价值专家

腾讯云架构师技术同盟

腾讯云创作之星

腾讯云开发者先锋

腾讯云代码助手

云原生构建

TAPD 敏捷项目管理

Cloud Studio

SDK中心

API中心

命令行工具

涵盖代码开发、场景应用、自动测试全流程，助你从零构建专属AI助手

一站式MCP教程库，解锁AI应用新玩法

我们目前正在设计我们的新数据库服务器，并且已经想出了一个折衷方案--我不完全确定如何回答。这些是我们的选择:48 or 1333 are，或96 or 1066 are。我的想法是，对于数据库服务器来说，RAM应该是充足的(我们有大量的数据，还有一些非常大的查询)，而不是尽可能快。显然，我们不能得到在1333兆赫的16 at芯片，因此上面的选择。那么，我们应该获得更多更慢的RAM，还是更慢的RAM

问数据库服务器:小快RAM还是大慢RAM？
EN

附加信息：

回答 7

Server Fault用户

Server Fault用户

Server Fault用户

社区

活动

圈层

关于

腾讯云开发者

热门产品

热门推荐

更多推荐

问数据库服务器:小快RAM还是大慢RAM？EN

附加信息：

回答 7

Server Fault用户

Server Fault用户

Server Fault用户

社区

活动

圈层

关于

腾讯云开发者

热门产品

热门推荐

更多推荐

问数据库服务器:小快RAM还是大慢RAM？
EN