Notes

Data flow is: application -> cache -> database (i.e. cache sits in between the application and the database). There are 3 strategies: Write through: write to both in parallel. Write around: database only. Write back: cache only. It basically depends on whether, during a write, the application writes to the underlying database, cache or both.

Four Seasons Hotel in Buenos Aires. Posadas 1086, C1011 CABA, Argentina

435: Hwy 527 & 180th St SE - near Fusion India. Olive Way & 6th Ave 412: Good starting positions: Puget Park Dr & Silver Firs Dr - seems residential but 5 mins away. 132nd St SE & 35th Ave SE. Or Amazon Go. Olive Way & 6th Ave

Just add something like the following after the last (useful) command in the shell script: curl -fsS -m 10 --retry 5 -o /dev/null https://hc-ping.com/.../$? Documentation here.

My life is feeling haphazard right now. There are too many small things in my head - I also have a lot of time on my plate these days, because I’ve taken some time off work, and a lot of these small things are just fillers for the free time I have on hand. Here are a few things, in no particular order (of importance etc.): What book do I start now? Finish creating a trust. Find a new team at work. What more on self-hosting front? Start preparing for interviews. Finish home setup. Start working out regularly.

Here is an easy way to setup a Wireguard VPN server, Pi-hole & Unbound on a Raspberry Pi. Note: Set private-domain: "ketanvijayvargiya.com" in unbound.conf file and restart the container. (In Portainer, it was located in the /var/lib/docker/volumes/wireguard-pihole-unbound_unbound_config/_data directory.) version: "3.8" services: unbound: image: mvance/unbound-rpi:latest restart: unless-stopped volumes: - unbound_config:/opt/unbound/etc/unbound/ networks: wg-easy: ipv4_address: 10.8.1.4 wg-easy: depends_on: [unbound, pihole] environment: - WG_HOST=... - PASSWORD= - WG_DEFAULT_DNS=10.8.1.3 - WG_DEFAULT_ADDRESS=10.8.0.x image: weejewel/wg-easy volumes: - wg-easy:/etc/wireguard ports: - "51820:51820/udp" - "51821:51821/tcp" restart: unless-stopped cap_add: - NET_ADMIN - SYS_MODULE sysctls: - net.ipv4.ip_forward=1 - net.ipv4.conf.all.src_valid_mark=1 networks: wg-easy: ipv4_address: 10.8.1.2 pihole: depends_on: [unbound] image: pihole/pihole dns: - 127.0.0.1 environment: - TZ=America/Los_Angeles - WEBPASSWORD= - ServerIP=10.8.1.3 - PIHOLE_DNS_=10.8.1.4 - DNSSEC=true volumes: - pihole_config:/etc/pihole - pihole_dnsmasq:/etc/dnsmasq.d ports: - "9001:80/tcp" cap_add: - NET_ADMIN restart: unless-stopped networks: wg-easy: ipv4_address: 10.8.1.3 networks: wg-easy: ipam: config: - subnet: 10.8.1.0/24 volumes: pihole_config: pihole_dnsmasq: wg-easy: unbound_config:

Key takeaways: Don’t work for money, otherwise you are only making your bosses richer. Try to create a side business. Understand the difference between liabilities and assets, then try to accrue more of the latter. Learn to invest. Take bold risks. Learn tax optimization strategies so you can keep more of what you earn. Work to learn - leave specialization for PhDs.

Key takeaways: It’s incorrect to assume that big & successful companies don’t innovate or their managers are incompetent & lazy. However, they still fail. Such companies innovate to make their existing products better. For e.g., improve sound quality of giant radios that sit in their own cabinets in living rooms of rich people. They even innovate to create new products, just that they fail to capitalize on these innovations. Because these products are initially crappy, the companies don’t release them to end customers. For e.g., small-sized radios that can fit in pockets. Scrappy startups are less stringent about the quality of the products they release, so iterate faster. By the time companies realize profitability of new products, they are already too late. Dilemma for big companies: they can’t invest in every dumb new idea. However, if they only invest in what works for them right now and wait for others to realize the profitability of new ideas, it’ll be too late by the time the latter happens. I think the proposed solution is to build startups within big companies, give them the autonomy to operate as they see fit and allow them to fail.

High level ideas: How to sleep is just another thing that kids need to learn. They learn things fast, with trial and error, so why should this be any different? They may struggle a bit but that’s how they learn other things as well. All kids, and even humans, wake up at night. The bigger concern is what they do after that. They cry because the environment isn’t the same as what put them to sleep. For example, mommy soothed them before sleep, so they require that again.

As usual, there are various things where I’d like to devote my time: Learn stuff by consuming content: this can be either through system design videos or arbitrary content on distributed systems. The idea is that this content consumption will (hopefully) make me smarter and help me get into a good company later on. Write code to learn things: the top contender for this idea is the labs from the distributed systems course from MIT. It’ll be uncomfortable and I’ll learn fundamentals that few people around me (at work) know about. However, I am slightly skeptical if I’ll ever be able to reuse this knowledge. (This is one of those things where you can only connect the dots looking backwards.) That Fast AI course on deep learning – but I know that won’t help me in any way in the future. Other ideas on consuming content: audiobooks, books etc. I have this fascination of learning new things that I think isn’t right. Is learning is my only goal, then #2 is the right approach. However, if my goal is to move to a better company AT&T some point in the future, #2 won’t be useful. ...