From time to time, we all get tired of the more formal side of work. When I do, I find myself musing on the wit and wisdom of my elders and betters. Below, are a few brief observations that I have acquired over the years attending RIR, IETF and the various NOG meetings worldwide.
Good ones are short: all the short ones are taken
Not to be confused with the album by Ian Hunter -this refers to DNS names. The reason this one is so good is that it’s immediately true: good DNS names are short and therefore automatically memorable, and because we now have so many people interested in a name in .com, all the two, three, four and five letter combinations of English language words (and some other strings) are now gone. In fact, most of them went before 2013. Does this help explain why so many people now want to register new gTLD names? Possibly it does!
RAID is not a backup solution
This was coined in response to the inevitable: people who previously took a backup on independent media, eg removable tape, or disks, thought that the Redundant Array of Inexpensive Disks (RAID) protected them against loss of media and forgot a couple of important things: 1) You sometimes accidentally delete things you want back, and when it’s removed from a RAID array it’s gone and; 2) Even RAID arrays can fail. I’ve suffered both kinds of failure, and had reason to wish I’d remembered that having it twice is better than having it once, reliably.
A little knowledge is a dangerous thing
Oftentimes people refer (disparagingly) to a “low clue zone” or “low clue density” and there is even a theory that the total amount of “clue” in the universe is a constant, like entropy and the second law of thermodynamics. If clue gets spread around too thinly to be useful then it’s wasted. What we need is more clue, but in concentrated places! Taken to its logical (reductionist) extreme perhaps this explains why Systems and Network admins are so reluctant to explain what they are doing…
Never attribute to malice that which is adequately explained by stupidity
This has been attributed variously to Goethe and Napoleon. But, usefully, crops up time and time again, especially in discussions around BGP and routing leaks. Almost all the events in recent times relating to breakdowns in global routing are misconfigurations, compounded by other misconfigurations (or lack of configurations) and shouldn’t be mistaken for deliberate attacks on the integrity of routing. It is a worry when the primary function of the RPKI system is to protect against malicious attacks on routing, and yet most of the breakdowns in BGP come from trusted individuals “inside the camp” mistakenly propagating bad information.
The good thing about standards is that there are so many to choose from
Attributed to Andrew Tanenbaum, this succinctly observes that the more we try and define things, the more likely it is we wind up with more than one definition to chose from. International standards are best when there are few of them applicable to any given situation. The IETF tries very hard to keep to one, but all you have to do is have one standard with choices (or, in IETF terms the magic words MAY or even SHOULD, because SHOUTY UPPERCASE words become the so-called “normative“, which take on definitional qualities) and suddenly, you really have two (or more) standards as people choose to implement the choices differently, in incompatible ways.
The DNS is not a directory
It’s amazing how many people want this not to be true, but as pointed out time and time again, and very nicely written up here in CircleID by John Levine, that’s not how it works. It’s not how the DNS is designed to work. Wanting it to behave like a directory isn’t the same thing as it being one. If you want a directory, stick to search engines.
So: what’s your favourite aphorism? Got a good one? Remember, good ones are short…
The views expressed by the authors of this blog are their own and do not necessarily reflect the views of APNIC. Please note a Code of Conduct applies to this blog.