What do you name your servers?
-
@IRJ Boring.
-
@anonymous said:
@IRJ Boring.
I was only joking. Alot of people like to name their servers by their function, but I like to use a theme. Themes are fun because you get to name each server and you hide their function from someone that only knows the server name.
-
We use a code like this....
[datacenter]-[os]-[application or function][number]
So a Toronto based Linux server for MySQL might be...
to-lnx-maria1
-
We use a ton of datacenters so we have a lot of DC codes. We have one for each Rackspace, Azure, Amazon, Digital Ocean and Vultr DC plus a few for our own. Very handy to be able to just look and see, instantly, where a workload exists.
-
@scottalanmiller when are you going to write up your Jumpbox How-To?
-
[business acronym][city, if relevant][function/software][number]
CSIVictoriaUtility01
PISEPrintServ01
PSFVExchange01
-
@scottalanmiller said:
We use a ton of datacenters so we have a lot of DC codes. We have one for each Rackspace, Azure, Amazon, Digital Ocean and Vultr DC plus a few for our own. Very handy to be able to just look and see, instantly, where a workload exists.
You can still have DC codes and fun names... rsp-w2012-Coruscant, do-lnx-Skywalker, etc... 8-)
Arguably, naming by function does make more sense when you have many servers, lol.
-
@anonymous said:
@scottalanmiller when are you going to write up your Jumpbox How-To?
And for reference, our real world jump box is...
dny-lnx-jump
-
@dafyre said:
Arguably, naming by function does make more sense when you have many servers, lol.
Back in the day, NTG used a naming convention of Austrian cities for our servers. This is circa 1999 - 2001.
Vienna - application server
Salzburg - database server
Graz - email and collaboration
Linz - virtualizationBut that didn't last too long. Once we started to grow it was obvious that it just made things a mess. Although it is amazing that I still remember Vienna and Salzburg like they were my children!
-
Vienna and Salzburg were Compaq Proliant 800s with Pentium III 500 MHz 100Mhz FSB. Each had 4x 9GB SCSI drives. Only one of the two had hardware RAID. Both were RAID 5. I believe that they each had 128MB of RAM. The ran NT 4 and both lasted months short of ten years without a failure.
-
@scottalanmiller Just using linux seems to vague, why not use distro instead?
-
@scottalanmiller said:
Vienna and Salzburg were Compaq Proliant 800s with Pentium III 500 MHz 100Mhz FSB. Each had 4x 9GB SCSI drives. Only one of the two had hardware RAID. Both were RAID 5. I believe that they each had 128MB of RAM. The ran NT 4 and both lasted months short of ten years without a failure.
They sure don't make them like they used to!
I remember pretty much every device I touched at my last job. Some of them I was glad to retire... others made me sad.
-
@scottalanmiller said:
Vienna and Salzburg were Compaq Proliant 800s with Pentium III 500 MHz 100Mhz FSB. Each had 4x 9GB SCSI drives. Only one of the two had hardware RAID. Both were RAID 5. I believe that they each had 128MB of RAM. The ran NT 4 and both lasted months short of ten years without a failure.
Mmmmm quantum hard drives - I remember the sound those make like it was yesterday. Did they have a sub-brand on those, like the desktop Fireball line? Was it Atlas? Good memories
-
@anonymous said:
@scottalanmiller Just using linux seems to vague, why not use distro instead?
How often do you need to really know that at a glance? What is more important is that the Linux teams know where to log in and which team needs to look at the box. It isn't like the Ubuntu team and the RHEL team and the Suse team have different people. But Linux and Windows do.
It's not a super amount of info, just enough for basic identification when needed quickly and to avoid errors.
-
Pinging dny-lnx-jump.ntg.co [65.75.137.152] with 32 bytes of data:
Request timed out.
Request timed out.
Request timed out.
Request timed out.Ping statistics for 65.75.137.152:
Packets: Sent = 4, Received = 0, Lost = 4 (100% loss),You have ping disabled?
-
You could have something like nyc-ub1404-mysql55-1
But then do you change the name when you update? How much details goes into a hostname? How long does it get?
-
@anonymous said:
Pinging dny-lnx-jump.ntg.co [65.75.137.152] with 32 bytes of data:
Request timed out.
Request timed out.
Request timed out.
Request timed out.Ping statistics for 65.75.137.152:
Packets: Sent = 4, Received = 0, Lost = 4 (100% loss),You have ping disabled?
We don't put our internal names into public DNS. That would be silly
-
@scottalanmiller said:
You could have something like nyc-ub1404-mysql55-1
But then do you change the name when you update? How much details goes into a hostname? How long does it get?
I didn't say anything about version numbers..... nyc-cent-web-1
-
@scottalanmiller said:
How often do you need to really know that at a glance?
You don't need to know at a glance, but why not? If you going to take up characters to define it as linux, why not give the distro instead?
-
@anonymous said:
@scottalanmiller said:
How often do you need to really know that at a glance?
You don't need to know at a glance, but why not? If you going to take up characters to define it as linux, why not give the distro instead?
Well if you are going to script things super quickly, it's nice to say...
for i in $(grep lnx servers); do ssh $i uptime; done
