GCP Snapshot restore

3 posts / 0 new
Last post
#1 Sat, 11/16/2019 - 16:57

GCP Snapshot restore

Hi Everyone, a very happy new user of Virtualmin here after having had a terrible experience with a hosting reseller. I decided to take the time to teach myself what it takes to manage shared hosting in a VPS.

Everything has fallen nicely into place except one critical component of my set-up... backup/restore.

I built the VPS in Google Cloud Platform, and set up a snapshot schedule to take an incremental backup of the entire boot disk (only one disk in this VPS) in the middle of the night each night. Those get created fine, and I'm able to create a new VPS instance using those snapshots. In order to test my backup procedures, I restored my most recent snapshot to the boot disk of a new VM with the exact same specs as my production machine, and edited my client host file to test the functioning of my hosted domains. However, while the VPS is booting up, it never finishes the boot sequence and throws hundreds of errors that all seem to point to not being able to reach the public internet, socket refusals, etc. (entire output of serial console included below).


I have a gut feeling that the machine can't recover unless it's on the configured external IP address - which can't be tested without taking the server offline temporarily. I wanted to check here to see if that is the likely cause before doing so, since I have live sites on there. I'm also worried that a server reboot right now might result in not being able to boot back up (since a snapshot of that same drive can't boot on a new VM).

Any insight from those experts lurking here would be very very appreciated!

Thanks in advance, and excited to join the Virtualmin users community!

Sun, 11/17/2019 - 03:15

Did you do a websearch for the several error messages?

Did you Guidelines read about the versions you are using?

(further these warnings is a misconfigure of postfix nothing todo with your problem :

Nov 16 17:42:34 node1 postmulti[1282]: /usr/sbin/postconf: warning: /etc/postfix/master.cf: undefined parameter: mua_client_restrictions
Nov 16 17:42:34 node1 postmulti[1282]: /usr/sbin/postconf: warning: /etc/postfix/master.cf: undefined parameter: mua_helo_restrictions
Nov 16 17:42:34 node1 postmulti[1282]: /usr/sbin/postconf: warning: /etc/postfix/master.cf: undefined parameter: mua_sender_restrictions
Nov 16 17:42:34 node1 postmulti[1282]: /usr/sbin/postconf: warning: /etc/postfix/master.cf: undefined parameter: mua_client_restrictions
Nov 16 17:42:34 node1 postmulti[1282]: /usr/sbin/postconf: warning: /etc/postfix/master.cf: undefined parameter: mua_helo_restrictions


The KEYS with phasfrase you are using with it are not ok / found?

That and this one maybe:

Nov 16 17:42:36 node1 systemd[1]: Started LSB: start and stop Stackdriver Agent.
Nov 16 17:42:36 node1 stackdriver-agent[847]:  * The instance has neither the application default credentials file nor the correct monitoring scopes; Exiting.
Nov 16 17:42:36 node1 stackdriver-agent[847]:  * Not starting, configuration/credentials error.
Nov 16 17:42:36 node1 stackdriver-agent[847]:    ...fail!

Thsi one to but if stackdriver not working ok.. Nov 16 17:43:15 node1 snapd[553]: stateengine.go:150: state ensure error: persistent network error: Get https://api.snapcraft.io/api/v1/snaps/assertions/snap-declaration/16/MJbt3BgxESyOON7gqKVEnA06NLRM3Dxd?max-format=3: dial tcp: lookup api.snapcraft.io: Temporary failure in name resolution

So search the web for those.

Post your versions here?

And sorry i can't help with those. , but versions for others who maybe can would be nice and is a part of forum guide here.

Don't also know what the cloudmin version do with this parts if using other ip's , and maybe if you solved some of above errors, still the network part / same ip's are needed, i didn't find / know a way to change such in virtualmin.

For bind there are scripst on the web to change if not doing manualy, but for networking and sockets, hmm

Sun, 11/17/2019 - 04:46

the snapshot should be an exact copy of the original. from some of the errors i am seeing, your ip address seems like a solution.

i see a mention of a bug that once existed in virtualmin about those postfix errors you are getting. not sure if its relevant or not...https://www.virtualmin.com/node/59160

I am also seeing a lot of temporary name resolution errors...sounds like this might be an ipaddress issue. Perhaps there are 2 choices here... 1. accept the snapshot works (without temporarily stopping your live one and spinning this one up using that static ip address) 2. stop the production system, fire up this one and see what happens. If the snapshot is a recent one, i dont think your clients sites would be down for long if its working as expected. Trouble is, if dns TTLs are long, their sites might not be a way of testing its working until TTLs expire.

AJECreative is the home of $5 webhosting, $15/month VPS servers (1cpu,1gb RAM, 25GB storage)
Centos7, Debian9, or Ubuntu18LTS
Available Control Panels = Centos-Webpanel, Cyberpanel, or Virtualmin


Topic locked