This morning I woke up to the sound of silence in my small server room. It appears the server was completely shut down, no fans spinning, no buzzing and no humming. A couple of days ago I was fixing various corruption issues and thought I did not fix it properly. It is Saturday, I want to work on my next woodworking project, but now I had to troubleshoot ANOTHER server issue. If this week was not enough already...

I turned on the server, then went to my workstation to try and connect to it. First small win - the server boots properly. So I fired up Event Viewer and filtered on errors and warnings. None were found - curious, I thought. I removed the filter and scrolled through the list - eventually finding this jewel:

Log Name:      System
Source:        Microsoft-Windows-Kernel-Power
Date:          2019-11-01 6:17:32 PM
Event ID:      109
Task Category: (103)
Level:         Information
Keywords:      (70368744177664),(1024),(4)
User:          SYSTEM
Computer:      SERVER
Description:
The kernel power manager has initiated a shutdown transition.

Shutdown Reason: Button or Lid

Yes, you heard right - it is 2019 and I STILL do not have any way to OCR an image natively in macOS Catalina. And no, Notes sucks. I do not want to OCR and index images, I want to convert images to text I can copy and paste and work with. To the best of my knowledge Notes only OCRs and indexes content to make it searchable, it does not allow you to copy text.

OneNote's OCR sucks. I'll add an image and after 30 minutes there is still no option to copy the text. Sometimes it takes up to an hour. That makes it a no starter.

It is 2019 and I assumed by now we will have native OCR support in our operating systems. So disappointed. At least I found this, but I should not have to.

I have one side table which I always have to wage war with my little one for real estate, so I decided to make my own side table instead of buying one. I did not spend hours designing it - it is really basic, made from Fir and finished with Polyurethane to maximize water resistance, as I will be placing various types of liquid on it.

Here it is...

So recently I detailed an ongoing issue I had with Telus. I did not check the traffic usage on Telus' website before the end of my billing cycle unfortunately, but I do know I checked about a week before the end and it was around 500GB, just 40GB more than my counters which is correct due to the changes I made in the beginning of the cycle.

So it is no surprise that I was shocked to get ANOTHER bill for $45 overage (472GB over 1TB):

I wanted to achieve a very simple thing - my graph showing CPU usage per core should be normalized. Regardless whether a machine has 8 cores or 2, I want to see the CPU usage as a percentage out of 100. This cannot be achieved using InfluxDB and InfluxQL at the moment as it does not support cross joining of measurements from InfluxDB as sourced from Telegraf, which stores the number of CPUs in the system measurement and the per core CPU usage in the cpu measurement.

Flux is a new language to be part of InfluxDB 2.0, which is currently still in Alpha. A more limited version of Flux is available with InfluxDB 1.7+. It took me quite some time but eventually I figured out how to write a Flux query that will work in Grafana with the BETA Flux (InfluxDB) plugin to create this: