Repository navigation
Constant "Couldn't get bundle details. Trying again after 10 seconds #119
|
So, I found I was having a capacity issue on my standalone host. I had two different LUN's and needed to combine them. New-HoloDeckInstance -Version "9.0.2.0" -InstanceID gray -vSANMode ESA -NsxEdgeClusterMgmtDomain -DeployVcfAutomation -DepotType "Online" Everything remained the same on the Holo Router, all of the binaries are there for 9.0.2. However, now it keeps giving the following: 28-04-2026 22:54:49 SetHoloDeckConfig[2283]: [INFO] Updating the Global Config, ConfigKey: vSANMode, ConfigValue: ESA I have tried stopping and restarting, but get the same results. Anyone else encountered or have a fix? |
Replies: 27 comments 2 replies
|
When you log in to VCF Installer and navigate to the depot page, do you see the bundle details populated? If not, right click on the browser > Inspect element > Network tab and look for the sddcs API call and see if VCF Installer is unable to get the bundle meta data. |
|
What about accessing the VCF Installer UI from webtop? |
|
Same situation here... I was able to download by SDDC manager but the Holodeck still has the same message |
|
Resolved |
|
Implemented the workaround. Works great. Thank you! |
|
What's the workaround you applied? Changed the -eq 7 to -ge 7? That would download the patch file as well causing this issue. Better way to solve this is to change "9.0.2.0.*" to "^9.0.(0|2).0.\d+$". Then you will have to delete the instance and deploy again because your current installer already has the bundles downloaded including the patch bundles |
|
On the post for bug 122, it mentions the following: Edit the module on the deployed HoloRouter: Kill the in-flight pwsh (the running session holds the unpatched function in memory): Restart the deploy from a fresh pwsh: So, to make sure I understand what you are saying it to delete what is partially deployed and start new with this command? Correct? |
|
No. Instead of doing this: sed -i 's/$vcf9_bundle_details.count -eq 7/$vcf9_bundle_details.count -ge 7/' Change "9.0.2.0.*" to "^9.0.(0|2).0.\d+$" in SddcMgmtDeployment.psm1 Then run: Remove-HoloDeckInstance -ResetHoloRouter |
|
My bad Change any occurrences of "9.0.2.0.*" to "^9.0.2.0.\d+$" including the one you showed in the screenshot |
|
So, I have apparently not edited something correctly. I am getting this error now. Here are the locations of the file edits. PS /root> cd /root/.local/share/powershell/Modules/HoloDeck/Modules/ |
|
Looks like you changed a lot more than needed. Revert back the changes and only update the following lines in SddcMgmtDeployment.psm1: 890: 894: 900: 993: 997: 1003: |
|
Can you log in to VCF Installer and share a screenshot from the depot page to see if the bundle details are populating in the UI? |
|
yes, click on depot settings and see if the binary details are displayed |
|
okay. Once the download is complete, you can open the state file using the following command in holorouter: Look for the json block "Install-VcfInstallerBundles" and change its status from "InProgress" to "Success". Save the file and run your New-HoloDeckInstance command and it should skip the binary download and continue deployment |
|
I did have another issue where it was hanging while deploying the automation and host 1 was memory constrained. Put it in maintenance mode and increased the memory from 128G to 132G. That moved past that issue, but then hit the above. So, I am not sure if having to interrupt the process caused this issue or not. |
|
Could be. You can check the deployment logs to know more |














Looks like you're hitting #122