BenQuick.me – Things about stuff

DNS and DHCP for the homelab

I keep threatening to set up a homelab. I want to keep it away from the devices that I already have on the network, and I want it to ‘just work’. That means I need a reliable DNS and DHCP setup.

Firstly I set up a new VLAN and allocated an IP address range to it. I’ve gone with VLAN 4 and 172.16.4.0/24. My Operating System of choice is FreeBSD with the latest supported release being 14.1, so I downloaded the 14.1 bootable ISO, mounted it in my virtual machine and began the install.

I didn’t do anything exciting during the install. Once the install had completed, I did a quick freebsd-update fetch followed by a freebsd-update install to get the system up to date, then it was time to go.

I looked in to what DHCP and DNS servers to install, and kept coming back to ISC BIND and ISC DHCP. They’re both tried and tested and simply work. I did look at PowerDNS, but I couldn’t get it to behave in the way I wanted.

Ooh, I got solar!

Back in November 2023, I got solar panels installed along with a ‘solar’ battery – I don’t like the term ‘solar battery’ as it’s not a thing. The battery isn’t solar, and doesn’t require solar. It’s literally a battery.

I’m UK based, so I used a solar calculator from the Energy Saving Trust to see if my roof would be viable for panels – I did see a roof around where I live that faced the same way as mine, but they only have panels on one side of their house. I wanted to see if I could have more panels installed.

It turned out that my roof is South-East and North-West facing, both are suitable for having solar panels installed on them. However, even though I specifically asked, most quotes I got didn’t accomodate for panels on the North-West side of the roof, which was frustrating. In fact, only two installers listened to what it was that I wanted.

Accessability no longer matters. Phew

The pandemic is, thankfully, over. We all need to return to our lives as they were pre-December 2019. At least I think that’s what various governments are telling us.

As are various vendors. Do you want to attend an event that they put on during the pandemic? Great! It’s in-person only once more.

Who needs accessabilty, right?

Can’t travel to an event? Sorry, pandemic’s over!

Can’t get childcare? Sorry, pandemic’s over!

Have hearing difficulties so need real-time captions? Sorry, pandemic’s over!

Can only dip in and out so need to be able to watch recorded sessions? Sorry, pandemic’s over!

It’s quite frustrating. Somewhat understandably, vendors were scrambling to be as inclusive as possible during the pandemic. They need to tell you what products they can sell you, so of course need to get that message out however they can. But now? Sorry pal, got to be there in person.

The difference in vendor demos

I’ve attended no end of vendor demos over the years, and some stick in your mind for different reasons – A NetApp demo at a University campus. This was a great demo with a good account team, but the use of the word ‘tetris’ does start to bug you after a while. A Packeteer demo on site where I worked, where the tech tried to tell us that their product could reduce the latency between Europe and North America. No amount of explaining that you simply cannot change physics like that would convince them otherwise

An unnamed vendor who, no matter how many times you re-phrased the question, were unable (or unwilling) to answer how their storage arrays replicated data 🚩

How to win friends and influence people

Sometimes I’m in a position to influence decisions. I normally base my opinion on the technical detail of a solution, and how it fits current and future business needs. But recently, I’ve realised how much the approach of the sales team comes in to play.

For example, Pure Storage are a joy to work with. They have a very strong product set which very much fits the needs of the business I currently work at. They’re easy to engage with as a company and an account team. We have approached them previously to ask for a solution to a problem, and they have been honest in excluding themselves from the running as the products they have weren’t a good fit. Wonderful. I believe in credit where it’s due, and am receiving nothing for saying this.

Then comes Rubrik. I have never contacted them. In the past 9 months, they have e-mailed me 18 times. 9 times in the last month alone. They have offered to pay me just to talk to them. That feels shady – Your product should speak for itself. How good is your product if you have to pay people to listen to you?

Sometimes I’m in a position to influence a decision, someimes I’m not If I’m ever in a position to influence vendor selection for any product Rubrik sell, I’ll try my hardest to ensure they’re not in the running. You’re never the only vendor available.

KPIs and SLAs – More important than outcomes?

Some time ago I was working in a vSphere role and I got an escalation through about an unexepcted reboot of a host. Sure, we all encounter a PSOD if we’re unlucky. They’re certainly not a normal occurace, and it’s a stop screen where the default config doesn’t result in a reboot. I started looking at the logs, vpxd.log, hostd.log, vpa.log, and so on. There was no indication of any failure. The logs just stopped, then re-started when the host was booting up.

That’s neither normal nor expected. No problem though, keep looking and something will show up. I had a look at the host SEL to see if anything showed up. The host had been rebooted by a user in a different team. Problem solved.

Why are we slowly heading back to inaccessibility?

When the pandemic hit businesses, understandably, panicked. Suddenly the majority of the workforce was remote, and that was a new thing to a lot of companies. How do you manage staff when they’re not sat at a desk you own? How do you know people are working if you can’t see them?

Zoom, WebEx, Teams enter stage left. For me, Teams has been a revolution – You can have a text chat with someone, turn it in to a phone call, a video call, a screen share with no real effort. Wonderful. Collaboration should be this easy, and in 2020 it was.

Industry events that used to be in-person only were suddenly, and hastily, thrown online. Yes, thrown. It’s different to put on an event that is online than one that is held in-person, and a lot of the first events I attended didn’t have time to re-assess how to deliver their content. Those events were painful to be honest, but it was new so nobody is really at fault.

There were a lot of clunky handovers from speaker to speaker, all waiting for the another to give the signal they had finished speaking / it was your turn to speak. There was a surprisingly large number of backgrounds that featured guitars, too.

Do we want to solve problems, or say we’ve solved problems?

The mystery plug in my office has piqued my interest. pic.twitter.com/mDwFJoXcWu
— Donal Coffey (@DonalkCoffey) October 5, 2021

I saw the above tweet the other day. It’s fairly amusing, and later I saw a few imply it was the process solution implemented as a result of the Facebook outage on 4th October 2021.

It resonated – How often do we implement solutions without solving problems? How often is a process introduced (‘Do not unplug’) that makes no attempt to solve the underlying problem (Whatever cannot be unplugged).

All too often we focus on implementing a solution without taking a step back to ensure it actually solves the problem. If we ignore all the bad stuff we can only talk about the good stuff, right? A report that’s all green is better than a report that has red on it, right? All that matters is the report, right?

Ostrich management never ends well.

Reducing power usage part 2

Since posting part 1 of reducing power, I received the order of my ASRock and have re-used the hard drives from the previous Shuttle I had. The Shuttle wasn’t loud by any stretch, but the ASRock is incredibly quiet. To the point where I have to check it’s still running!

After previously being impressed with how long my hard drives have lasted, one went pop during a power failure. That’s a bit disappointing, but the drives are only Western Digital Blue’s which are just generic consumer desktop disks.

Continue reading Reducing power usage part 2

Reducing power usage part 1

A good 7 years ago, I bought a Shuttle SX38P2 Pro which has been running pretty much 24/7 ever since. Recently the PSU fan has been struggling, but I can’t complain for the time it’s been in use. Over those 7ish years I’ve suffered a single drive failure.

# smartctl -a /dev/ada0 | grep Power_On_Hours 9 Power_On_Hours 0x0032 081 081 000 Old_age Always - 13972 # smartctl -a /dev/ada1 | grep Power_On_Hours 9 Power_On_Hours 0x0032 015 015 000 Old_age Always - 62738

As the fan was struggling, I contacted Shuttle to ask about replacements but received no reply. That was disappointing, but I guess understandable as the model has been discontinued (I didn’t think the cooling was model-specific though).

Continue reading Reducing power usage part 1