I'm trying to do a V1000 install and it's not working. The only thing I'm doing different than all my other installs is that we have to set the IP manually instead of through DHCP. It's having an issue during boot right before the log-in prompt:
File system 'vol0' successfully created.
Select FSM "none"
Starting FSS locally.
Start FSS "vol0"
FSS 'vol0' start initiated.
vol0[0] pid 7088
State: LAUNCHED (no substate) Datestamp
Last Admin: START Datestamp
Launches 1, core dumps 0, flags <!CONFIGURED,AUTO-START>
Select FSM "none"
Cannot activate FSS vol0
Activate FSM "vol0"
mount.cvfs: Can't mount filesystem 'vol0' No such process
System log may contain additional details.
Could not mount cvfs filesystem. Aborting...
Failed to create SNFS correctly, aborting firstboot!
I don't know the password to log-in to the console and I've never had this happen before. Any suggestions? We've already tried deleting the appliance and redeploying it. I can't deploy it on an interface with DHCP, even temporarily.
I got a copy of the V1000 User Manual and it seems that I have to enable the CLI accounts before I can use them. How do I do that with no GUI?
edit:
I was able to log in as cliadmin/cliadmin and am now at a bash prompt. Not all linux commands are working so I'm not sure what I can do from here now.
Tags:
Hi Mark,
All of those settings should be set as part of the Appliance import, shouldn't it? We never even looked at the client settings so we didn't change them. ESXi 5.0 and the host is only running 2 other VMs that are not doing anything at the moment.
We did re-import a fresh image when we tried again. I'm in my lab now and will be seeing what I can reproduce here where I can share access if necessary.
Thank you,
Ray
Yes, all of those settings are in the meta file.
We still have seen customers run into issues if the system is highly overloaded.
In a nutshell, this isn't the DHCP issue that I thought it was.
If indeed you can reproduce this issue on another system on the same network, that would be very interesting.
Hi Ray,
I deleted the former reply that I have provided because the information was not quite correct.
I just manually installed V1000 on a network with No DHCP and the installation proceeded as documented within the Quick Start Guide (step 17) where I was prompted for a manual IP address.
Please review the minimum hardware requirements and perhaps try again. It is possible that there could be an odd timing issue preventing the V1000 from booting as it should normally. Just make sure that you aren't running a broken DHCP (that is, DHCP that isn't serving up IP addresses) and you should be fine.
Oh, BTW, did you happen to have an external USB thumb drive on the ESX host at the time that you were booting the V1000, or perhaps mapped to the virtual machine? This is a stretch, BTW, just trying to rule out a few things.
Oh yes, and one more thing... is this an Intel or AMD processor box that you are using?
Per our data sheet, we only support Intel processors at this point. If this is AMD, then that might explain the error.
Hi Mark,
I'll have to check on the chipset used by the ESXi hosts tomorrow when I'm on site. When you say it's not supported do you know if that's because it causes problems, or because QTM is being cautious?
I have not been able to reproduce the problem in Keeper's lab space either so I'm going to suspend the VM and take it to the customer and see if we can get it running that way like you initially suggested. Also, will the IP address have to be set every time the VM is rebooted or just the first 2 times? The guide isn't clear on that.
Thank you,
Ray
Regarding AMD : the box will not boot and will fail with mounting / creating the StorNext file system - which is essentially the error you are seeing. It will not work with AMD processors.
Regarding manual IP address : You will supply the IP address once during initial boot, then manually one more time in the first reboot. After that, you will login to the GUI and set the IP address manually through the GUI and will NOT need to set it again. It is persistent after that point.
One other quick question : are you using VLAN tagging?
Currently, VLAN tagging doesn't work with static IP address for the DXi V1000. This will be addressed in a future update. Note that using DHCP on a VLAN tagged LAN works fine.
Just got on site and found out that they are using VLAN tagging. I have a copy of the appliance that I built in our lab like you suggested. Will this still not work with VLAN tagging? It's in the state where I booted it and let the FS's get built then shut it down so I could export it.
Is the issue just for the initial boot like we are seeing, or will it not work at all once past initial boot?
They are not using AMD hosts, all are Intel.
It may work now that you have the appliance built. Remember that to get the IP address to 'stick' you will need to get the system booted to the point that you can manually configure the IP address through the DXi GUI. This means you have proceeded through the steps where you have entered the licenses through the GUI.
I *think* that the issue is more about the initial boot. Please do try it out and let us know.
The method building the appliance on one machine and moving it to another seems to have worked. The VM is running and VMPro is talking to it.
Comparing the install process of VMPro and DXi the DXi method needs to be cleaned up and made more robust to be able to handle large enterprise VM environments. I don't know what the link is between the network and FS being able to be created, but that seems like a silly link to have. Maybe deploy the appliance with the FS already created instead of having it do that on first boot? Which is basically what I did to get around the issue.
Thank you for your help,
Ray
© 2024 Created by Quantum Forum V. Powered by