• Set up the splunk logging composition, for the most part, and then removed it.
      • Docker has logger drivers which tell it how to access the logs from the containers. The default is json-file.
      • Splunk uses HTTP Event Collection as the logging driver, known as HEC.
      • You can run 500Mb/day through Splunk for free.
      • Created an account with splunk.com.
      • Remember that it takes a minute or two for the splunk server to start. Check for status: healthy in docker-compose ps.
      • Added a token that the app container uses to push the http events to a specific url.
      • Installed the docker monitoring app inside the splunk interface.
      • You need collectord to make it work, you can’t just forward it directly from hec. https://www.outcoldsolutions.com/.
      • This is way more complicated than it needs to be. I literally just want to tail a logfile from the browser. The docker interface already forwards stdout/stderr from the container.
      • I also don’t want to bloat the under-resourced DO droplet with a splunkenterprise container. It’s built to do way more than I’d use it for.
    • Darnold had a passer rating of 3.6 last night. Insanely low. Patriots D are absolutely killing it, they’re in the top 10 for all offensive players in fantasy. I lost the yahoo league and won the espn league. Got 2/5 in supercontest.
    • In purging some docker images/containers on my laptop, it tried to upgrade the database container to 12 on a new build/pull. That version is incompatible with a db created with 11 (my version), so I pinned it in docker-compose.
    • Big Mouth roles I didn’t know: John Mulaney is Andrew, Chelsea Peretti is Nick’s phone, Kristen Bell is Jay’s pillow, and Jordan Peele is Duke Ellington!
    • Did blender research.
      • Notes.
        • I don’t need pulse. I only need it for smoothies (and maybe nut butters), I don’t use the blender for salsa, grinding herbs, chopping nuts, etc. Plus I still have the ninja if need be.
        • I don’t need preprogrammed settings. Soups, desserts, smoothies, blah.
        • I don’t need digital or touch control. Buttons/switches are fine.
        • I don’t need self-detect and self-clean. Manual control is fine.
        • I don’t need timers and programmability. It’s not an oven.
        • I don’t need an app to pair with it wirelessly (lol).
      • Vitamix.
        • Given all the above, the 5200 is the right fit.
        • There are a couple more features to consider.
          • The classic series blenders have 2hp motors, while the smart series blenders have 2.2hp motors.
          • The smart series blenders are a bit smaller (low profile 64oz container).
          • The smart series blenders are a bit quieter.
        • Those 3 features aren’t worth it to me.
        • Its usual price is $450, there was one available for $400. The base is red, but worth the $50 discount. Ordered.
    • Camelcamelcamel is a great price tracker for buying expensive items on amazon. You can see the most/least expensive it has been, and even set up alerts where it will email you.
    • Bought a huge batch of shelled, skinned, blanched, unsalted peanuts. They’re much cheaper in bulk than other nuts – down as low as $2/lb. I bought 10lbs at $3/lb. At costco, the packages of pecans/cashews/etc are $8/lb, and that’s a great price already.
    • Downloaded paypal for my phone.
    • Greg bought week 9 for the sbsc banner. Franky has week 8. Jcriss paid but still needs to send the graphic, so he’ll probably get week 10.
    • Netflix issued 2b of corporate bonds to fund new content, adding debt.
    • WeWork is in a bad place.
    • FF waivers. 3 for yahoo, 2 for espn.
    • Used the new razor. Cleanup was easy, since it catches all the hair. Cleaning the unit is easy, just run under the tap (full or taken apart). It’s a close shave. I need to get used to the motion/pressure so it gets everything.
    • Updated the banner_archive (.gitignored):
      • 2019-01-niners.svg
      • 2019-02-warriors.svg
      • 2019-03-williamstown_basketball.svg
      • 2019-04-moon_landing.txt
      • 2019-05-meats.jpg
      • 2019-06-ducati.jpg
      • 2019-07-britney_spears.txt
    • Made the static/README.md always show build status successful for the stack view, since the gitlab project is private. The regular readme still points to the actual build badge, updating in realtime.
    • Started the new season (9) of AHS. It’s called 1984, and it’s an 80s camp horror theme (which I’m a big fan of).
    • Would be nice if spotify had a simple feature like “save for later” where you could add artists (or albums, or songs) to a list to go through whenever you wanted. I often find a few new artists, and then find a couple more good suggestions in “related artists”, but there’s no way to keep those artists queued in the app to go through later.
    • My laptop battery:
      • Discharge, normal use, full brightness: ~12.5%/hr. Total time ~8hrs. Linear.
      • Discharge, idle, closed: ~0.25%/hr. ~2% over 8hrs. Linear.
      • Charge, normal use, full brightness: ~50%/hr during the linear portion to ~85%SOC. Total time ~3hrs. Logarithmic toward the end.
    • Google Keep can do bulleted lists by default, just type an asterisk then space on a new line.
      • Moved a few longer term items from blogger to google keep.
    • The new batch of protein bars is even better than the previous, which is saying a lot.
    • Made all my gitlab projects private, and transferred them all from my user namespace to my group namespace. Updated all remotes, on my laptop and the droplet.
    • Updated the readme in the fivesquared project to shrink the images to a reasonable size. They now display side-by-side.
    • Al Noor lunch with Corey/Harner.
    • Tradebot.
      • Created and configured the repository in gitlab.
      • Enabled autodevops at first to play with it. It created a pipeline with 3 jobs: build, code_quality, and test. If would have added a deployment job as well, but you need a kubernetes cluster. The first job, build, failed and stopped. The default command was `herokuish buildpack build`. I don’t need autodevops for this project, so I disabled it.
      • Added a very basic .gitlab-ci.yml pipeline instead. Just runs pylint via tox.
      • Created both tickets, one for analysis and the other for buy/sell.
    • Useful advice: when performing any sort of corrective action (filing a complaint, etc) against a party in a non-face-to-face manner, pretend you’re a third-party witness. You won’t be discounted as disgruntled, and they’ll perceive more weight around the incident if it’s apparently large enough to be reported by a witness.
    • LFS for prod db backups and the banner archive.
      • curl -s https://packagecloud.io/install/repositories/github/git-lfs/script.deb.sh | sudo bash
      • sudo apt-get install git-lfs=2.9.0
      • cd ~/code/supercontest
      • git lfs install
      • git lfs track “backups/
      • git lfs track “supercontest/static/img/banner_archive/
      • git add .gitattributes
      • Then commit and push as usual.
      • You can run `git lfs status` to see which files are going to git and which are going to lfs.
    • Added the sbsc production db backups to git lfs. Added the banner_archive as well.
    • Listened to some good old Tremonti solo work today. Lead vocals/guitar for creed/alterbridge.
    • GitLab has template files for .gitignore, .gitlab-ci.yml, LICENSE, and Dockerfile that you can add to any project. Example: https://gitlab.com/bmahlstedt-group/tradebot/new/master. They will come prepopulated with whatever environment syntax you want (Python, for most my cases).
    • Ducati is owned by Lamborghini, which is owned by Audi, which is owned by Volkswagen.
    • Went through some different lists of folks’ “stocks you’ll never sell”. Curated them into ones I actually cared about. Removed ones that didn’t look amazing (unless your basis was 10 years ago). Removed the ones that I don’t believe in (general electric GE, etc). Removed the ones that I don’t want to support (mcdonalds MCD, etc). The remainder is obvious. The big players with consistently linear positive trends. After some recession to reduce the position, these would be good ones to dump some excess into (or weight more highly in the tradebot, as a qualitative boost since it’s a quantitative tool).
      • AMZN
      • GOOG
      • MSFT
      • DIS (disney)
      • IDXX (idexx laboratories: products and services for animal food and stuff)
      • ILMN (illumina, inc: genetics)
      • BAM (brookfield asset management: equity)
      • RACE (ferrari)
      • BA (boeing)
      • LMT (lockheed martin)
      • INTC (intel)
      • CSCO (cisco)