Tezos主网络节点：硬件规格

• CPU:没什么特别的,可以达到几个GHz,tezos支持 多核井
• 记忆:请参阅下面的内容.似乎> 4GB是合理的. 8GB应该足够了.
• 存储:随机读取和随机写入> 150MB/s.这很容易成为嵌入式系统的瓶颈,不建议使用旋转磁盘.闪存卡也可能磨损 如果它们不是工业级产品,则很快.

• CPU: nothing special, a couple of GHz will do, tezos supports multi-core well
• MEMORY: see my plot below. Seems like >4GB makes sense. 8GB should be enough.
• STORAGE: Random read and random write >150MB/s. This can easily become a bottleneck with embedded systems, a spinning disk is not recommended. Also flash cards could wear out quickly if they are not industrial grade.

建议8GB RAM ¹ 表示一个节点.Kiln建议使用10GB的RAM.我的Ubuntu 18.04 Baker/Node中的mainnet节点不会占用CPU的大量资源.但是,强烈建议使用SSD.至少我在HDD上运行的第二个节点上出现了很多"超时"错误.SSD上运行的主网节点运行平稳.

8GB RAM is recommended¹ for a node. Kiln recommends 10GB of RAM for example. The mainnet node in my Ubuntu 18.04 Baker/Node doesn't use up a lot of resources from the CPU. However, using an SSD is highly recommended. I at least have been getting a lot of "timed out" errors on the second node I run on an HDD. The mainnet node running on an SSD runs smooth.

我们在Digital Ocean上运行我们的节点.我们的节点是2vCPU/2GB RAM/60GB SSD,具有4GB交换空间.我们已经将此配置运行了6个月以上,没有任何问题.我们曾经以1vCPU运行,但是有很多iowait.第二个vCPU解决了该问题.SSD几乎是必不可少的.目前,超过4GB RAM或超过2个CPU/内核的一切都不过分.也许在1-2年以上,当有更多智能合约在运行时,您将需要额外的CPU容量.

We run our node on Digital Ocean. Our nodes are 2vCPU/2GB RAM/60GB SSD with 4GB swap. We have been running this configuration for 6+ months with no issues. We used to run with 1vCPU, but had lots of iowait. A second vCPU solved that. SSD is almost a must-have. Anything over >4GB RAM or >2 CPUs/Cores is just overkill at this time. Maybe in 1-2+ years when there are more smart contracts running you'll need the additional CPU capacity.