ESXi 6.7 Troubleshooting
-
This is fine for a lab but do not do this for production. It could make upgrading/reloading much more difficult later.
For VMware, I would use USB or SD cards if possible. ESXI is only like 500 MB's and that is it. You can store the ESXi logs on normal storage. There are numerous sites out there with instructions.
Wow, I had no idea. We've been doing it wrong for over a year now :man_facepalming_medium_skin_tone:
-
@DustinB3403 said in ESXi 6.7 Troubleshooting:
ESXi recommends SD card, USB is a bit to fragile for the number of writes you'd be making to it.
I agree with @DustinB3403 My new 440 has SD cards in a RAID 1 so I feel much better.
On the flip side, I have only had four USB keys go bad on me in 11 years with 13 servers. It is not a big sample size but it is something.
I like to use SanDisk Ultra Fit 32GB. Going strong for 6 years so far. Plus they are small so it can won't mess with the bezel or knock it with your hand or shoulder. Transcend kinda suck. Had three of those go bad on me over 8 years.
-
@pmoncho how do you deal with USB or SD card failures in this instance? Is there a way to program a replacement or would it be as easy as re-downloading ESXi onto another USB/SD and inserting that as a replacement?
-
....Mirroring aside
-
You can use these instructions or others you find on google.
https://kb.vmware.com/s/article/2042141So, if the USB dies, you reload the same version of ESXi onto new USB key, keeping the storage drive in tact, set IP address and then restore the config and reboot. Boom! all done. It is that easy as I had to do it once.
I recommend backing up the config every time you make a change. I never really delved into profiles either but from what I hear, you can save those in vCenter. Once you add the server back, reload the profile, reboot and you should be good to go.
ESXi is loaded all into memory after boot up. Other than USB disk checks, nothing is stored on the USB key after the hypervisor is loaded. It mostly writes to the USB key on reboot/shutdown.
If you use a USB/SD key, you will need to create a log folder on your storage volume to save all the log files. After install, set the ScratchConfig parameter with the correct path and reboot (It is easy so don't worry).
https://kb.vmware.com/s/article/1033696If you want to really go nuts on DR. Backup the config while it is loading, take out the USB key, use Linux (I use clonezilla) and clone the USB key. Then put the key back in the server. You should be good.
-
@pmoncho said in ESXi 6.7 Troubleshooting:
On the flip side, I have only had four USB keys go bad on me in 11 years with 13 servers. It is not a big sample size but it is something.
WOW - that's huge! nearly 30% failure...
-
@pmoncho said in ESXi 6.7 Troubleshooting:
If you want to really go nuts on DR. Backup the config while it is loading, take out the USB key, use Linux (I use clonezilla) and clone the USB key. Then put the key back in the server. You should be good.
Did you mean loaded, as in online and working?
Why do it then and not just shut it down, pull it and make a copy?
-
@Dashrender said in ESXi 6.7 Troubleshooting:
@pmoncho said in ESXi 6.7 Troubleshooting:
If you want to really go nuts on DR. Backup the config while it is loading, take out the USB key, use Linux (I use clonezilla) and clone the USB key. Then put the key back in the server. You should be good.
Did you mean loaded, as in online and working?
Why do it then and not just shut it down, pull it and make a copy?
Oops my bad. Yes I did. typing in haste as I was headed out the door.
One can shut it down and then make a copy but I have done it while the system was up and running also.
-
@Dashrender said in ESXi 6.7 Troubleshooting:
@pmoncho said in ESXi 6.7 Troubleshooting:
On the flip side, I have only had four USB keys go bad on me in 11 years with 13 servers. It is not a big sample size but it is something.
WOW - that's huge! nearly 30% failure...
Yeah. The previous admin used shitty USB keys. Once I switched to Sandisk, I have not had any problems. One was and old USB key with a microcenter label, the other three were Transcend. Man, those SUCK!
-
@G-I-Jones said in ESXi 6.7 Troubleshooting:
@pmoncho how do you deal with USB or SD card failures in this instance? Is there a way to program a replacement or would it be as easy as re-downloading ESXi onto another USB/SD and inserting that as a replacement?
Does your Dell have the SD card reader? If so, it runs the SD cards as a RAID1 array.
I haven't installed ESXi myself ever (to expensive too quickly once you get into "I have to license this"). I'd want to use 2 usb drives in a mirror if going the usb route, but don't know if that is easy or possible with the installer.
-
@travisdh1 it doesn’t. I know if you order them specifically for ESXi then they ship it with either an SD or USB (can’t remember which) that’s already setup as RAID 1. Definitely a detail to remember for the future. Pushing hard to go Virtual after this last Ransomeware hit.
-
Here's a KB that matches the virtual machine hardware versions with product versions like you see when you try to create a VM - https://kb.vmware.com/s/article/1003746.
Here's a list of capabilities that match to different virtual machine hardware versions - https://docs.vmware.com/en/VMware-vSphere/6.7/com.vmware.vsphere.vm_admin.doc/GUID-789C3913-1053-4850-A0F0-E29C3D32B6DA.html.
So if you created a VM that was compatible with 6.7U2, it could not run on a host running any version of ESXi older than 6.7U2. In your case that won't be a problem since you are just getting started, but it is something to keep in mind if you build an environment with hosts on newer versions of ESXi later and want to move things around. Keep the VM hardware version at the right level for max flexibility / compatibility in the environment.
-
Also, just for due diligence, I recommend checking the power profile on the server and ensure it is set to max performance. Turn off processor C-states as well to ensure you get all the firepower that host can give you. There's a host power power management policy setting in ESXi for each host that you can set to high performance as well.
-
@NetworkNerd, @pmoncho , I'm still having no luck over here. I've tried 6.5, 6.7 and 6.7 U2. All with Dell ESXi and without. Dell is saying that they think it's because this specific server doesn't have a PERC cache. Does that sound like nonsense to you all or not? I know the previous machine (which I had 0 zero issues other than forgetting to install VMWare Tools on initially), worked fine, but it DOES have a PERC cache...
-
@G-I-Jones said in ESXi 6.7 Troubleshooting:
@NetworkNerd, @pmoncho , I'm still having no luck over here. I've tried 6.5, 6.7 and 6.7 U2. All with Dell ESXi and without. Dell is saying that they think it's because this specific server doesn't have a PERC cache. Does that sound like nonsense to you all or not? I know the previous machine (which I had 0 zero issues other than forgetting to install VMWare Tools on initially), worked fine, but it DOES have a PERC cache...
Which RAID card do you have in the server according to the BIOS or iDRAC?
-
@pmoncho said in ESXi 6.7 Troubleshooting:
@G-I-Jones said in ESXi 6.7 Troubleshooting:
@NetworkNerd, @pmoncho , I'm still having no luck over here. I've tried 6.5, 6.7 and 6.7 U2. All with Dell ESXi and without. Dell is saying that they think it's because this specific server doesn't have a PERC cache. Does that sound like nonsense to you all or not? I know the previous machine (which I had 0 zero issues other than forgetting to install VMWare Tools on initially), worked fine, but it DOES have a PERC cache...
Which RAID card do you have in the server according to the BIOS or iDRAC?
PERC H330 Adapter
-
@G-I-Jones said in ESXi 6.7 Troubleshooting:
@pmoncho said in ESXi 6.7 Troubleshooting:
@G-I-Jones said in ESXi 6.7 Troubleshooting:
@NetworkNerd, @pmoncho , I'm still having no luck over here. I've tried 6.5, 6.7 and 6.7 U2. All with Dell ESXi and without. Dell is saying that they think it's because this specific server doesn't have a PERC cache. Does that sound like nonsense to you all or not? I know the previous machine (which I had 0 zero issues other than forgetting to install VMWare Tools on initially), worked fine, but it DOES have a PERC cache...
Which RAID card do you have in the server according to the BIOS or iDRAC?
PERC H330 Adapter
Well, they are right that the PERC H330 does not have a cache. Based on a few google searches, others have had similar issues with this card depending on the RAID level.
So, what type of drives do you have and what RAID level are you using?
-
@pmoncho said in ESXi 6.7 Troubleshooting:
@G-I-Jones said in ESXi 6.7 Troubleshooting:
@pmoncho said in ESXi 6.7 Troubleshooting:
@G-I-Jones said in ESXi 6.7 Troubleshooting:
@NetworkNerd, @pmoncho , I'm still having no luck over here. I've tried 6.5, 6.7 and 6.7 U2. All with Dell ESXi and without. Dell is saying that they think it's because this specific server doesn't have a PERC cache. Does that sound like nonsense to you all or not? I know the previous machine (which I had 0 zero issues other than forgetting to install VMWare Tools on initially), worked fine, but it DOES have a PERC cache...
Which RAID card do you have in the server according to the BIOS or iDRAC?
PERC H330 Adapter
Well, they are right that the PERC H330 does not have a cache. Based on a few google searches, others have had similar issues with this card depending on the RAID level.
So, what type of drives do you have and what RAID level are you using?
RAID 5, 7 - 1TB Intel SSD, shipped with the Server.
-
@pmoncho, @NetworkNerd, @travisdh1, @Dashrender, @DustinB3403, @marcinozga - Thanks for your help. After reading an article on it or two and talking with Dell, We're ditching the H330. Boss just ordered the H730P.
-
@G-I-Jones said in ESXi 6.7 Troubleshooting:
@pmoncho We're ditching the H330. Boss just ordered the H730P.
That, in the long run, is probably the best call. Any way you can get the boss to add on the SD Card module? That will be a benefit also if you stay with ESXi.
