Tonight we had a discussion with Steve Pope, Solarflare’s CTO and Founder, on Solarflare’s new “Firewall in the NIC” capability called ServerLock that goes into Beta this quarter. During our time together we reviewed the following:
What exactly does Solarflare mean when they say they’ve put a firewall in the NIC?
How does this improve the security of my server?
What is micro-segmentation, and how can this be applied to my applications, containers or VMs?
Why is having a firewall in the NIC better than a software firewall which is part of the OS?
Why is a firewall in the NIC better than say a top of rack firewall?
How much might this cost me in latency if the NIC is filtering every packet?
Who has Solarflare built this product for?
Where in my enterprise infrastructure should I consider using such a NIC?
Can this be used as an edge solution to enhance the security of my customer facing web servers, possibly further protecting them from a DDoS attack?
Tonight we had a discussion with Vahan Sardaryan, CEO & Founder of LDA Technologies, to talk about their new STAC-T0 Benchmark which set the bar surprisingly low, 98 nanoseconds, for network tick to trade latency. During that call, we reviewed the following:
What is the significance of 98 nanoseconds?
Who worked with LDA to make this possible?
In layman’s terms, how was the feat accomplished?
The jitter of the solution was six to nine nanoseconds.
What is the significance of having STAC validate this achievement?
This was months worth of work, what sort of ghost hunting was required?
Who could benefit from this technological advance?
The report highlights a measurement of -68 nanoseconds, can we trade into the future?
Where do we go from here, and can the bar be set even lower?
Scott would like to apologize for the quality of this podcast as both him and Vahan had to dial into the recording system due to, you guessed it, networking issues.
On September 14, Bob Van Valzah dropped in for a long session on Digital Currencies which turned into a three-part series on Blockchains. One or more series will follow covering all the issues surrounding digital currencies. To the left is a shot of Bob’s power bill for his Bitcoin Mining Rig that we discussed tonight, and also a video of that Mining Rig in production.
In the second part of our six-part series on Digital Currencies we’ll address the following questions around Blockchains:
So what are miners actually doing?
The enormous numbers, for example, the 10^18 power behind solving these mining problems.
Are we looking for primes in a huge field of numbers (hint, no)? Solutions have 130 zeros at the end or 2^130
Mining rigs as space heaters, the power issue.
Who decides which miner is a winner in extending the blockchain?
What happens when two miners each find a valid solution during the time of the flood?
Is it still one bitcoin block every 10 minutes?
What happens if it takes longer than 10 minutes to arrive at a solution?
What keeps the bad guys from circumventing the whole mining system?
How often would you win a bitcoin, and what is a mining pool?
In part three we’ll finish describing what is a Blockchain. While this has already been recorded and it should be released a few days following part 2.
Last night, September 14, Bob Van Valzah dropped in for a long session on Digital Currencies which turned into a three-part series on Blockchains. One or more series will follow covering all the issues surrounding digital currencies. To the left is a shot of Bob’s Bitcoin Mining Rig that we discussed tonight, and also a video of that Mining Rig in production.
In the first part of our six-part series on Digital Currencies we’ll address the following questions around Blockchains:
How did you get into bitcoin mining?
What is a Blockchain?
How is a Blockchain structured?
Where are Blockchains stored?
What is the role of a node?
Are all nodes miners?
Is a miner a person or a computer?
What is a mining rig?
In part two we’ll cover mining itself. This has already been recorded and it should be released a few days after part 1.
On August 8th, 2017 Luca Deri the founder of ntop was kind enough to call in and discuss Network Performance Monitoring (NPM). Luca is a well-recognized leader in the field of NPM having started at IBM Research back in 1993. He then went on to launch ntop in 1997 and has never looked back. In this specific market PF_Ring, n2Disk and several other tools by Luca and his team have become the de facto standard by which all others are compared. Today we discussed NPM then and now, touching on the abuse of port 80 for a wide range of wrapped traffic. We then moved on to packet capture, and the need for smart NICs moving forward if we wish to keep pace with faster networks like 50 and 100GbE.