Node Launching taking 5 minutes

Community Support VM Configuration / Contextualization
1

Hi,
Here is the node launch details

Tue Feb 4 16:12:17 2025 [Z0][VM][I]: New state is ACTIVE
Tue Feb 4 16:12:17 2025 [Z0][VM][I]: New LCM state is PROLOG
Tue Feb 4 16:13:21 2025 [Z0][VM][I]: New LCM state is BOOT
Tue Feb 4 16:13:21 2025 [Z0][VMM][I]: Generating deployment file: /var/lib/one/vms/228561/deployment.0
Tue Feb 4 16:17:53 2025 [Z0][LCM][I]: VM reported RUNNING by the drivers
Tue Feb 4 16:17:53 2025 [Z0][VM][I]: New LCM state is RUNNING

Its taking nearly 5 minutes to come up the node with the image size of 15gb.

2

Hi @mac,

You have asked a similar question here: Opennebula too slow

Without giving us much hints or details, could you please ellaborate a bit more?

Thank you,

3

Hi,

I have configured as follows
One Frontend server
One DB Server
One Image server withmounted with Frontend server

Have added 60 hosts into the Opennebula for deploying the VMs (auto deployement)

Each vm takes 5 minutes to deploye it.

Whereas in AWS while i am doing an testing the VM deployed in 30 seconds and the node is coming up over the network (tested with ping)

4

Do you have any error log?
With this is complicated to guess what could be, since it could be:

  1. Resource Constraints in terms of the CPU, or storage resources.
  2. Issues with the connectivity between the server and the node. It might be a packet loss, for example.
  3. Issues with the Storage backend.
  4. Hypervisor Configuration problem
  5. Is the AWS instance configured the same?, does it have 60 hosts as well?

Also, are you running the latest version of OpenNebula?
Please check /var/log/one/oned.log (as we already asked you on the other topic), and paste the information.

Thanks,

5

log (36.9 KB)

please find the logs of /var/log/one/oned.log

6

Any update on this

7

Hello @mac,

Sorry for the late reply, but the log contains several errors, and it looks like the environment you are using needs some preparation in advance. Probably your AWS environment has these pre-configured somehow?

For issues with scalability, I would suggest to check our documentation: Scalability & troubleshooting

On the log, we can see:

  1. VM is locked (those are at least on the first 5 lines): You will need to run: onevm show <VM_ID>, to check which VMs are locked and consider them to unlock.
  2. Slow SQL queries, so If using MySQL/MariaDB, ensure performance tuning with proper indexing. You’ll probably need to check the status, and then run OPTIMIZE TABLE history, vm_pool; to optimize tables.
  3. Host monitoring errors, meaning that OpenNebula is unable to retrieve monitoring data from these hosts, possibly due to network issues, host failures, or a misconfiguration. You will probably need to restart the monitoring by systemctl restart opennebula
  4. Other bottlenecks, probably you’ll need to configure the scheduler. Go to the specific config file (/etc/one/sched.conf) and check if it’s optimized:
  • Increase MAX_DISPATCH if VMs are queued for too long.
  • Reduce SCHED_INTERVAL if scheduling delays are an issue.
    (More info on scheduler on our documentation page).

Hope it helps.