My Rambling Thoughts

The Day The World Blue Screened

Happened to me at 12:34 pm. After my notebook power-cycled, a colleague told me his notebook also crashed, but unlike mine, it could not boot. He said it happened to many people in office.

A solution was posted on Reddit very quickly — enter Recovery mode and either delete the offending file or rename the entire CrowdStrike dir.

But there is a gotcha: to enter Recovery mode, you need the Recovery key if your system is encrypted with BitLocker. Who has it written down? Panic is when the server holding the keys is also BSOD. Pray IT has the key to the server offline...

I renamed the CrowdStrike dir and it was gone. I wanted a less intrusive change by using RegEdit to stop the service from running, but I could not do so. I had no admin rights to edit the Registry even in Safe mode.

It struck me I could disable ZScaler the same way... :devil:

In Reddit, there were postings that it affected N-1 and N-2 deployments too. N-1 means one patch earlier, N-2 means two. Generally servers will use N-1. To use N-2 means to be extra-conservative.

The reason this affected all is because the update is "data", and data is rolled out to all immediately. CrowdStrike prides itself having extremely up-to-minute protection. Data is normally considered safe — except when it is corrupted.

The lessons here:

  • Input validation!
  • Staged rollout (limits damage)
  • Roll back on error

Some people will say more testing before release. But we all know how that works. Testing is never comprehensive.

Against Trump no matter the cost

When you can't win legally, you take the other path.

Trump is very lucky to be alive.

TDS (Trump Derangement Syndrome) is real. Many people instinctly dislike him because they see him as rich, spoiled, a playboy and is not a career politician.

Let me put it this way. If the State Apparatus dug all the way to 40 years ago and could only bring such insignificant and weak charges against Trump, I'll say he has lived a pretty clean and good life.

And this is why the Deep State is worried... he cannot be blackmailed.

Conspiracy theory time

The gunman was seen climbing up the roof. Several bystanders notified SS (Secret Service) officers, but they were ignored.

The gunman was well within SS "secure" perimeter.

The SS sniper already had the gunman in his sight, but only fired after the gunman fired 8 shots.

The conspiracy theory is: kill the gunman after he kills Trump to leave no living witness. (This is not the first time this has happened, seriously.)

Taking guesswork out of electrical usage


Tuya Smart Plug

Get the Tuya Smart Plug with Power Monitor feature.

First thing first. This is an IoT device and requires Internet connection (phones home to Tuya Cloud). Its security is suspect. By default, it is controlled using a mobile app. It can be controlled remotely.

(Tuya devices can be controlled without Internet access, but it requires some setup.)

The 20A model can be gotten at S$7+ on Lazada. 16A should be ample — the UK plug is rated 13A only. I don't see how it can be used to measure 20A appliances.

It consumes some power. It uses 15W/day in on state, <5W/day in off state. If you go wild and make all your appliances "smart", you will add several kWh usage every month! (e.g. an extra 9.2 kWh/month for 20 appliances.)

To get accurate power usage, the plug should be kept on (up to an hour after last use). It takes time to upload readings to Cloud. It is not real-time. Even so, these devices are considered "chatty".

It does not show devices drawing <3.5W, but it still accumulates their usage.

Devices that use <0.5W draw 10W/day. I have 10 to 15 such devices around. Collectively, they use 4.6 kWh per month.

Appliances with soft power button still draw power when off. Appliances with hard power button may be truly powered off.

The fridge uses ~2 kWh daily. This is ~65% my estimate, but is twice its rated power consumption (0.98 kWh/day).

After measuring several devices in soft-off or sleep mode, my takeaway is:

  • If it uses <0.5W, can keep it on 24/7 (0.37 kWh/month)
  • If it uses >1.5W (1.1 kWh/month) and is rarely used, e.g. once every few days, it is better to turn it off

Power usage by appliance

AppliancePower (W)Est usageMthly kWh
Air-con 1149010h/day223.5
Dehumidifiers2008h/day48.8
Dryer21002h/week16.8
Fridge 226024/795.2
Hair dryer16005m/day4.1
Iron240030m/mth1.2
Kettle220016m/day17.9
Kitchen hood?00
Microwave12103m/week0.3
Oven?00
Toaster9804m/day2.0
Vacuum cleaner 3105060m/mth1.1
Washing machine 420004h/week3.2
Water dispenser 5200024/7?
Water heater450015m/day34.3
 
Fans10018h/day54.9
Lights4712h/day17.1
Air purifiers1124/78.1
Air quality monitors1.2524/73.7
 
PC2224/716.1
Notebook1324/79.5
Modem4.224/73.1
Routers1124/78.1
Chargers156h/day5.5

1 With one cooling unit at 50% duty cycle

2 I'm assuming the fridge uses 130W to maintain coldness

3 It takes a lot of power to create suction

4 When using warm or hot water. I estimate it uses 200W with cold water

5 It is in sleep mode when not dispensing water

It takes a lot of energy to produce instant heat.

The top 3 energy consumers are air-con (223.5 kWh), fridge (95.2 kWh) and fans (54.9 kWh), followed closely by dehumidifiers (48.8 kWh).

Fans use a lot of energy, surprisingly.

How to calculate power usage

If an appliance uses 10W, it uses 10 Wh in one hour. It'll use 240 Wh in a day and 7.32 kWh in a month (if 24/7).

Another example. If an appliance uses 1000W but is only used for 10 minutes a day, it'll use 166.7 Wh in a day and 5.08 kWh in a month.

The electrical tariff is currently S$32.47 cents/kWh (including GST).

1 kWh is actually quite cheap. The trick is to use less of appliances that use >10 kWh per month.

Top energy guzzlers

Don't need to think. The top 3 are: air-con, fridge and water heater. The question is, how much do they consume?

Unfortunately, we are not able to measure the power consumption of air-con and water heater directly, so we can only estimate.

My fridge is rated to use 358 kWh/year (3 energy star rating). Is this true? In contrast, my dryer is rated at 683 kWh/year.

Estimating air-con

My air-con is MA56EV1 (compressor) with four FT25DVM FCUs (Fan Coil Unit). It is non-inverter.

#FCUUnit capacity (kW)Total power (kW)
12.561.49
22.151.52
31.781.60
41.451.64

(The full table is more complicated cos there can be a mix of 2.5, 3.5, 5.0 kW units.)

Running four 2.5 kW units just about max out the capacity (5.80 vs 5.94 kW).

It does not take much more energy to run 2 or 3 units.

Assuming one air-con is turned on for 10 hours a day and the compressor runs for 5 hours, the energy consumed is 223.5 kWh/month.

Estimating water heater

My water heater is Bennington C600. It is rated to use 4.5 to 5.3 kW. It requires 20A plug. A standard 13A plug allows max 3 kW.

Assuming 15 minutes usage daily, the energy consumed is 34.3 kWh/month.

Estimating lights

LED lights do not use much power, but they still add up over time.

I estimate I turn on 6 10W LED lights for 4 hours at night and 4 of them for 8 hours in the day — my living room is rather dark. It works out to be 17.1 kWh/month.

Estimating ceiling fan

I cannot find the power rating for my ceiling fan. I normally set it to lowish medium speed. My guess is 60W. It is kept on for a long time, sometimes even for the entire day. If it's 12 hours on average, it works out to be 22.0 kWh/month.