I have browsed this site on and off for years; I marvel at the latest twitter postings stating EPYC / Platinum reviews of zillion core, terrabyte ram and nvme raid rigs.
That's great and all if you are running a company the size of GE but what is lacking, is regular reviews of smaller use speedball rigs.
I run a server that handles a handful of users at any one time. They run reports that are generated in milliseconds, and sometimes seconds. There are regular updates to the databases each day, which take many minutes. In this instance a 128 core quad 8 way rig with terrabytes of ram is absolutely no use at all.
So what about some reviews of the latest Intel / AMD rigs that have the highest clock speeds? What about different types of ram settings for the best throughput and latency for mysql databases that are run in Linux ramdrives? What about articles on tuning Linux servers for such rigs; the O/S, hardware and CFLAG settings?
If you had to build a rig today, which is used to serve a handful of users at a time, who require data to be retrieved from a database, collated, prepared and served as fast as possible what technology would you use?
That's great and all if you are running a company the size of GE but what is lacking, is regular reviews of smaller use speedball rigs.
I run a server that handles a handful of users at any one time. They run reports that are generated in milliseconds, and sometimes seconds. There are regular updates to the databases each day, which take many minutes. In this instance a 128 core quad 8 way rig with terrabytes of ram is absolutely no use at all.
So what about some reviews of the latest Intel / AMD rigs that have the highest clock speeds? What about different types of ram settings for the best throughput and latency for mysql databases that are run in Linux ramdrives? What about articles on tuning Linux servers for such rigs; the O/S, hardware and CFLAG settings?
If you had to build a rig today, which is used to serve a handful of users at a time, who require data to be retrieved from a database, collated, prepared and served as fast as possible what technology would you use?