New toy
FWIW, Iāve been happy with Remmina for remote management.
Ive been using remmina for RDP connections from fedora for about a year now.
Also works with built-in macOS VNC which a lot of options donāt (performance is kind of crap though).
o>o oh, VNC
hooray context
Ha, yes. My bad
Serious discussion(s);
1.) What are the drawbacks to automation?
2.) What are the drawbacks to moving things to the cloud?
Hereās my crack at answering.
1.) One drawback I can see about automation would be that if you automate everything and someone comes in to replace you that didnāt construct the automation, fixing it when it breaks could be a major issue.
2.) One drawback to moving things to the cloud was stated by ESR ā if the cloud provider no longer wants to do business with you, they own the infrastructure and could force you out. The cloud is just mainframes other people own anyway. I also know about the 80/20 rule, and although delineated by the big fish, the little fish do it and donāt tell their customers.
To me :
It take longer to set up and manage for a one shoot, you have to way the benefit and how much to automate. And you have the risk to break everything at once if you donāt implement safety check.
When it break on the cloud provider side (like twice this years with GCP, and I was on call both time grumble) there is NOTHING you can do.
Having your infra mean you know your backup plan and how to pull the service back from the death with all the satanic ritual that it takeā¦
To me, no mater what, donāt stick to a single cloud provider, have a DRP on another platform
On top of the points already rasied:
- it can lead to issues if things are just homogenized that shouldnāt be. Not everything needs (or even wants) snapshot backups, thank you very much.
- if you donāt need the scaling the cloud offers itās often more expensive than just buying servers.
Thereās also the issue that you canāt just (legally) put everything in the cloud due to privacy and data protection laws, or even simply due to internal or contractual rules.
Youāre at their mercy.
It costs 3-5x what it would to buy the hardware and colo it yourself.
It means youāre learning a bunch of new tools per cloud provider.
Once you automate your job, we donāt need you.
Then 6 months after youāre fired and replaced by a helpdesk, AWS changes their API with zero notice and everything breaks.
In my prior job, I used to automate my own tasks and not tell anyoneā¦ Manager thought I was a workhorse, and I ended up with a lot of free time.
If your company doesnāt have itās head up itās ass, when you automate your job, theyāll give you a raise and tell you to automate everyone elseās job and then fire everyone but you.
That said, a head not up an ass is probably in the minority.
This is why you automate and donāt tell anyone about it.
edit:
mostly kidding, but if working for a head up ass company as per aboveā¦
Iāve got a whole private git repo full of things I totally didnāt automate.
This is urban legend, yeah? Iāve had two jobs where automation was king and expected, and they seemed to need us more than ever.
The list of infrastructure/I.T. tasks is endless.
Hmmā¦ Anyone experienced with this?
I thought this was related to AT&T. But now Iām on Digital Ocean and the exact same thing is happening.
I canāt find anything in any log: Nginx, dmesg, syslog, messagesā¦
I have no idea what this is. The only thing common is Nginx. This has happened across different applications (Go and Node), different distros (CentOS and Ubuntu), and different providers (my KVM and Digital Ocean).
it worked hereā¦ is it okay there?