Previously, our system exclusively used a shared datastore to store VMs. However, the previous admin told me that there were some problems and that he disabled it. Now we use a local ssh datastore for new VMs, however the old VMs are still around and still use the shared datastore (which is fine for these old VMs).
One day ago one user undeployed one of his old VMs (which still uses the old datastore). However, it could not successfully deploy and is now stuck in a PENDING state which I cannot resolve. First I thought it is because of space problems, however, I manually freed up some space on all hosts but it still won’t deploy. Upon further inspection in oned.log it turns out the deploy action fails continuously because:
one.vm.deploy result FAILURE [one.vm.deploy] Image Datastore does not support transfer mode: shared
To resolve the problems I tried to:
- enable the old datastore again, as I thought maybe it prevents the scheduler from selecting it, but it did not resolve the problem
- I went into the image datastore and set TM_MAD to “shared” via onedatastore update, however, that did also not do anything (of course I also tried to enable the datastore again with this change applied).
Image datastore output from onedatastore (with TM_MAD setting reverted):
shared datastore output:
Version used: 5.8.5
I did not restart one after the datastore change… could that be a problem?
Thanks in advance