I have a custom Linux VM that was migrated from a local Hyper-V instance. The VM is a virtual appliance running CentOS 5.1 that is provided to us from a vendor. I have limited abilities to customize, but I am able to install some RPM packages.
My first attempt to run this VM was without the WAAgent installed. The disk was able to be uploaded successfully, the VM was able to be created successfully but there was no communication outside of the Azure network segment (I was able to connected to the VM using a windows console on the same network segment).
After some research, I found that it seemed to be necessary to install the WAAgent in the VM. I found a precompiled version of the agent and installed the RPM and the prerequisites (python-pyasn1-0.0.12a-1.el5.rf.noarch.rpm, epel-release-5-4.noarch.rpm, and WALinuxAgent-2.0.8-1.noarch.rpm). When I uploaded the disk and createe the VM this time, all worked OK.
Fast forward a couple of weeks and yesterday I installed a new SSL certificate on the VM. All still seemed OK (meaning I had successful outside communication). I came in this morning to check the VM and I was unable to communicate with it again. After some troubleshooting (I was able to connect to the VM via SSH and HTTPS from the console server inside the Azure network segment, but I was still unable to communicate to the internet and vice versa, including DNS resolution from the Linux VM. I also rebooted the VM a few times to no avail. I also recreated the endpoints and it didn't make a difference).
The only thing that seems to resolve this is to delete the VM, delete the disk and reupload the disk and recreate the VM.
Anyone have any ideas? Thank you in advance!
Keith