Live data from Hacker News

RFC 1178: Choosing a name for your computer (1990)

tools.ietf.org

71–80 of 128 posts

Re: RFC 1178: Choosing a name for your computer (1990)

#71
Over 5 years at working at Amazon (quite some time ago, I don't know what they do now) and we happily violated "no domain names" without any issue for linux servers. If you don't have linux boxes that span different networks it doesn't cause a problem to have server have one single domain name and have the hostname be the same. Linux boxes aren't proper security devices or router so don't use them as that and don't mutihome them. The routers and switches were given short names. These RFCs simply don't capture that there are rules you can adopt (and if you're big enough -- which is not terribly big) then you can discard some of these suggestions completely. There is not a one-size-fits-all. I ignored this RFC's suggestion at a fairly massive scale for 5 years and never had an issue.

Also don't fall into the trap of just assigning random serial numbers to servers. I later worked somewhere that did this and it makes it difficult to communicate about the servers in the middle of outages. I've had communication issues where I've been talking to another engineer and I was using the first hex digits of the server name and they were using the last as shortcuts and we thought we were logged into different servers and it was the same one. You hardly ever want humans dealing with your server names, but when humans do need to use your server names it is one of the times that really matter because shit is on fire.

Group them by single purpose of what the cluster does with some kind of incrementing number. The idea to use theme names and not "project" names is also deeply 100% wrong. When you have 100,000s of servers you run out of theme names and you'll fail to remember the name schemes in the middle of an outage. Name them after what the servers do, and keep them more or less reflecting their purpose. Consider carefully some kind of numbering scheme to keep the short names unique across datacenters so you don't have a dozen foo-101 servers. You may want to use incrementing serial numbers for both datacenters and cluster members or something so "foo-1-101" and number your datacenters (or logical cluster number if you're really big and stamp them out 30,000 at a time or something).

Oh right this is the RFC from 1990. Yeah, shit has changed, this RFC needs to evolve.

Back in 1990 when this was written a single system admin hand managing 20-30 servers was a lot. Web didn't exist. I can't recall any kind of load balancing or much clustering. You might have SunOS boxes running RIP doing routing across internal subnets that were 10baseT. NAT and firewalls weren't used much at all and servers would just sit on public IPs. This RFC is prehistoric.

Re: RFC 1178: Choosing a name for your computer (1990)

#72

Over 5 years at working at Amazon (quite some time ago, I don't know what they do now) and we happily violated "no domain names" without any issue for linux servers. If you don't have linux boxes that span different networks it doesn't cause a problem to have server have one single domain name and have the hostname be the same. Linux boxes aren't proper security devices or router so don't use them as that and don't m…

> Oh right this is the RFC from 1990. Yeah, shit has changed, this RFC needs to evolve.

Like, don't name your MX wolfsschanze. It may cause problems with cancel culture raging on.

Re: RFC 1178: Choosing a name for your computer (1990)

#73
post #70

https://en.wikipedia.org/wiki/List_of_Greek_mythological_fig... has been good for naming my workstations for decades. Dual-boot scenaria can use Roman/Greek names for the Mac/Windows or Win/Linux “personalities” of the same physical machine.

I once almost got fired because my computer was called aphrodite, the CTO had to demo my webapp, and he couldn’t spell it correctly.

Re: RFC 1178: Choosing a name for your computer (1990)

#74

This kind of naming is fine for small projects, and personal setups. In professional settings I would suggest using descriptive names and rely more on DNS. It's not that I don't enjoy fun and personal naming schemes, it's just that it's a constant annoyance when dealing with a large number of different systems. We've been hired to deal with different companies, who picks some random naming, cars, athletes, plants, ci…

Cute naming schemes are a holdover from the days of physical servers when location, function, etc were subject to change so an immutable machine name couldn't contain them. Now that most things are virtual and easily created/destroyed there's no good reason _not_ to name functionally. The most common refrain I've heard in defense of these archaic naming practices is that "we don't want the hackers to know which serve…

> Now that most things are virtual and easily created/destroyed there's no good reason _not_ to name functionally.

Uh, my servers with at least 256G of RAM and a raid60 of more drives than you have virtual machines can't easily be moved around.

> If you are dealing with assets in a physical environment then obviously the above doesn't apply.

Yes, baremetal, the only way to guarantee performance!

Re: RFC 1178: Choosing a name for your computer (1990)

#75
post #63

I use Harry Potter character names. There were hundreds of the mentioned off hand in the books, so I usually use obscure ones. They have the benefit of being easy to pronounce and spell: Mafalda Hopkirk, Silvanus Kettleburn, Irma Pince, etc. https://en.m.wikipedia.org/wiki/List_of_Harry_Potter_charact...

All devices in our house are creatures: griphook, fawkes, crookshanks, pigwidgeon, hedwig, aragog, firenze, grawp, etc.

Nice! I usually do creatures for more appliance-type devices and wizards for servers. But it is somewhat of an arbitrary division.

Re: RFC 1178: Choosing a name for your computer (1990)

#76
Best naming scheme I ever saw was at Harbinger in the late nineties. They used the periodic system. The last octet of the IP address corresponded to the atomic number, and you could use the full element name, or the abbreviation. So carbon.harbinger.com was x.x.x.6 and c.harbinger.com was a CNAME to carbon. oxygen or o was .8 etcetera.

Re: RFC 1178: Choosing a name for your computer (1990)

#79
post #76

Best naming scheme I ever saw was at Harbinger in the late nineties. They used the periodic system. The last octet of the IP address corresponded to the atomic number, and you could use the full element name, or the abbreviation. So carbon.harbinger.com was x.x.x.6 and c.harbinger.com was a CNAME to carbon. oxygen or o was .8 etcetera.

When you split a workload from one to multiple machines, would it have to follow the rules of fission to decide where to run it?

Re: RFC 1178: Choosing a name for your computer (1990)

#80
post #21

Earlier quoted context omitted.

Why is percussion01 better than web01?

I'm assuming (based on reading how they performed grouping) that percussion01 didn't exist, but perhaps cymbal01 did (since a cymbal is a type of percussion instrument). Therefore cymbal01 would be a web server of sometime. The benefit to naming it this way is you could have multiple types of web servers (internal, dev, prod) and by using a more generic name you could more easily change the function of that web serve…

My feeling, though, is that you care more about the function of a machine than the physical machine itself. Why not change the machine name when you change it’s purpose? There is no reason you need to know that the current prod web server used to be the dev server.
Post reply on HN