Sir, I think it is always better not to put VM's in local datastore. If in real time one host crashes, and you have 5 vm's - two in local and three in shared. You can save three but how will you save the two ? Even if you bring vendor and replace with new hardware or ESXI host itself, you will re install ESXI and re configure everything. But you cannot restore those two VM's which were in local. One possibility is if you had taken any backup , you can restore , or else its gone!! SIr, in my office I told you there were three ESXI hosts, I haven't configured any clusters or HA. and many a times there was power outages for whole day and we had UPS batteries with 3hours backup and all the time all these hosts gets off after battery backup time. It restarts after power comes. So my question if I had configured HA, how it could be ? who will be master and who will be slaves? And most importantly Customer will face serious issue, because all VM's are down for whole day. And lets imagine another scenario if the three host gets corrupted together due to some natural calamity...
You configured HA over three hosts and due to power failure all three hosts are down so HA cannot do anything simple. You have to implement disaster recovery to overcome natural disasters.
Sir, When you tried to move vm6 to LUN 10 there was some issue. You said it might be an issue with Datastore cluster. So my question, to perform storage Vmotion it is not necessary to have Storage Cluster ? You meant you need to have shared storage, it will work Correct ? But after sometime , you tried moving LinuxS1 to LUN10 and it got failed eventhough there was no storage cluster, a bit confusing. Could you please clarrify ?
You don’t need to have storage cluster When the moving failed i think it was specifically an issue with datastore not cluster. He should have explained and tried to troubleshoot
Thank you . Not even paid classe gives us such deatiled information .. thank you
Thank you for your feedback!!
What will happen if the SAN storage is down? Is there any solution for that rather than Raid?
Sir your explanation s are very clear do you have VMware training videos I'm intrigued to purchase
Sir, I think it is always better not to put VM's in local datastore. If in real time one host crashes, and you have 5 vm's - two in local and three in shared. You can save three but how will you save the two ? Even if you bring vendor and replace with new hardware or ESXI host itself, you will re install ESXI and re configure everything. But you cannot restore those two VM's which were in local. One possibility is if you had taken any backup , you can restore , or else its gone!!
SIr, in my office I told you there were three ESXI hosts, I haven't configured any clusters or HA. and many a times there was power outages for whole day and we had UPS batteries with 3hours backup and all the time all these hosts gets off after battery backup time. It restarts after power comes. So my question if I had configured HA, how it could be ? who will be master and who will be slaves? And most importantly Customer will face serious issue, because all VM's are down for whole day. And lets imagine another scenario if the three host gets corrupted together due to some natural calamity...
You configured HA over three hosts and due to power failure all three hosts are down so HA cannot do anything simple. You have to implement disaster recovery to overcome natural disasters.
Hi sir,is thare any off line classes available
Not now
@@maverickcloud tnx for reply sir
Sir, When you tried to move vm6 to LUN 10 there was some issue. You said it might be an issue with Datastore cluster. So my question, to perform storage Vmotion it is not necessary to have Storage Cluster ? You meant you need to have shared storage, it will work Correct ? But after sometime , you tried moving LinuxS1 to LUN10 and it got failed eventhough there was no storage cluster, a bit confusing. Could you please clarrify ?
You don’t need to have storage cluster
When the moving failed i think it was specifically an issue with datastore not cluster. He should have explained and tried to troubleshoot