Early experience with using Linux tc to fight bufferbloat latency
Over on the Fediverse I mentioned something recently :
Current status: doing extremely "I don't know what I'm really doing,I'm copying from a website¹" things with Linux tc to see if I canimprove my home Internet latency under load without doing too muchdamage to bandwidth or breaking my firewall rules. So far, it seems towork and things² claim to like the result.
¹
² https://bufferbloat.libreqos.com/ via @davecb
Distribution source packages and whether or not to embed in the source code
When I described my current ideal Linux source package format , I said that it should be embedded inthe source code of the software being packaged. In a comment, bitprophet had a perfectly reasonableand good preference the other way:
Re: other points: all else equal I think I vaguely prefer the Arch"repo contains just the extras/instructions + a reference to theupstream source" approach as it's cleaner overall, and makes it easierto do "more often than it ought to be" cursed things
The Amanda backup system and "dump promotion"
The Amanda backup system is what we use to handle our backups. Oneof Amanda's core concepts is a 'dump cycle', the amount of timebetween normally scheduled full backups for filesystems. If youhave a dumpcycle of 7 days and Amanda does a full backup of afilesystem on Monday, its normal schedule for the next full backupis next Monday. However, Amanda can 'promote' a full backup aheadof schedule if it believes there's room for the full backup in agiven backup run
What 24 hours of traffic looks like to our main web server in January 2026
One of the services we operatefor the department is a traditionalApache-based shared web server, with things like people's home pages( eg ), pages for various groups,and so on (we call this our departmental web server). This webserver has been there for a very long time and its URLs have spreadeverywhere, and in the process it's become quite popular for somethings . Thesedays there are a lot of things crawling everything in sight, andour server has no general
Why we have some AC units on one of our our internal networks
I mentioned on the Fediverse a while back that we have airconditioners on our internal network . Well, technicallywhat we have on the internal network is separate (and optional)controller devices that connect to the physical AC units themselves,but as they say, this is close enough. Of course there's a storyhere :
Why do we have networked AC controllers? Well, they control portableAC units that are in our machine rooms for emergency use, and havingtheir controllers on our internal network means we can
Prometheus, Let's Encrypt, and making sure all our TLS certificates are monitored
I recently wrote about the complexities of getting programs toreport the TLS certificates they use ,where I theorized about writing a script to scrape this informationout of places like the Apache configuration files, and then today I realized the obvious specific approach for our environment :
Obvious realization is obvious: since we universally use Let'sEncrypt with certbot and follow standard naming, I can just look in/etc/letsencrypt/live to find all live TLS certificates and (a) hostname for them, for cross-checking
Some notes to myself on Super-based bindings in GNU Emacs
I recently had to deal with GNU Emacs lsp-mode in a context where Icared a bit about its keybindings ,and in the process of that ran across mention of what one could callits leader prefix, s-l. People who use GNU Emacs a lot will know whatthis specific 's-' notation means, but I'm not one of them, so it tookme a bit of research to work it out. This is GNU Emacs' notation for'Super', one of the
A small suggestion in modern Linux: take screenshots (before upgrades)
Mike Hoye recently wrote Powering Up , which isin part about helping people install (desktop) Linux, and theFediverse thread version of it reminded me of something that Idon't do enough of :
A related thing I've taken to doing before potential lurching changes(like Linux distribution upgrades) is to take screenshots and windowimages. Because comparing a now and then image is a heck of a loteasier than restoring backups, and I can look at it repeatedly as Ifix things on the new setup
The complexities of getting programs to report the TLS certificates they use
One of the practical reasons that TLS certificates have dangerousexpiry times is that in mostenvironments, it's up to you to remember to add monitoring for eachTLS certificate that you use, either as part of general purposemonitoring of the service orspecific monitoring for certificate expiry. It would be nice ifprograms that used TLS certificates inherently monitored theirexpiry , but that's a fairly big change (forexample, you have to decide how to send alerts about that information).A nominally easier change would be
A Go question: how do you test <code>select</code> based code?
A while back I wrote an entry about understanding reading allavailable things from a Go channel (with a timeout) , where the code used two
selectsto, well, let me quote myself:The goal of waitReadAll() is to either receive (read) all currentlyavailable items from a channel (possibly a buffered one) or to timeout if nothing shows up in time. This requires two nested selects,with the inner one in a for loop.
In a recent comment on that entry, Aristotle
Menu