Good evening,
I'm facing an issue wherein some of our clients refuse to PXE boot after having previously PXE booted in the past. Unfortunately, this is related to an air-gapped environment, so I can't share relevant screenshots, logs, etc. Primarily, I am running into this when testing out my task sequences. I am removing the computer from the MECM console in each iteration so that it is "fresh". To clarify, the lack of PXE booting affects the client BEFORE the NPB file is sent over. The SMS Pxe logs show that the client is (correctly) identified as an unknown computer and in state that a client reply is requested after the boot image is advertised to it. In spite of this, the client continues to wait for a PXE deployment that never comes.
With that said, just to run through some common things:
1 - All new workstations are fine. This only affects previously imaged workstations. This is notable in that it comes up when I am testing task sequences and the like.
2 - Deployments are available to all systems and all unknown devices in this environment. Unknown computer support is enabled. Duplicate hardware is not showing up in the Site heirarchy.
3 - The network is fine. There's plenty of DHCP space available, IP Helpers are assigned, and fresh boxes have no issue.
Now that that's out of the way, I'd like to mention the following:
The client is COMPLETELY removed from the MECM console. There is no associated with the SMS Guid or MAC address. That's in the MECM Console... When I inspected my database, I found the following table, and I suspect that it's what is causing me problems:
dbo.System_Disc
In this table, there is a GUID that has some random characters before showing the problem client's MAC address. It's something to the effect of 1234-12312312-ABCDEF123456. This repeats several times. There are other rows that are present as well, such as client0, Client_Type0, Unknown0, Active0, Obsolete0, Reassigned0, Decommissioned0, and Management0. The non-existant client (at least in the MECM console) has several rows within this system discovery table. I suspect that it's somehow interfering with MECM's ability to push out a valid PXE image. I reviewed the maintenance tasks in MECM, and one of them is "Delete Obsolete Client Discovery Data". Is there a way that this can be run manually instead of on a schedule? OR can I force removal of the client? I suspect our weird behavior is entirely related to the device being removed from the console, but still present in the database. Any help would be welcomed. Thank you!
I'm facing an issue wherein some of our clients refuse to PXE boot after having previously PXE booted in the past. Unfortunately, this is related to an air-gapped environment, so I can't share relevant screenshots, logs, etc. Primarily, I am running into this when testing out my task sequences. I am removing the computer from the MECM console in each iteration so that it is "fresh". To clarify, the lack of PXE booting affects the client BEFORE the NPB file is sent over. The SMS Pxe logs show that the client is (correctly) identified as an unknown computer and in state that a client reply is requested after the boot image is advertised to it. In spite of this, the client continues to wait for a PXE deployment that never comes.
With that said, just to run through some common things:
1 - All new workstations are fine. This only affects previously imaged workstations. This is notable in that it comes up when I am testing task sequences and the like.
2 - Deployments are available to all systems and all unknown devices in this environment. Unknown computer support is enabled. Duplicate hardware is not showing up in the Site heirarchy.
3 - The network is fine. There's plenty of DHCP space available, IP Helpers are assigned, and fresh boxes have no issue.
Now that that's out of the way, I'd like to mention the following:
The client is COMPLETELY removed from the MECM console. There is no associated with the SMS Guid or MAC address. That's in the MECM Console... When I inspected my database, I found the following table, and I suspect that it's what is causing me problems:
dbo.System_Disc
In this table, there is a GUID that has some random characters before showing the problem client's MAC address. It's something to the effect of 1234-12312312-ABCDEF123456. This repeats several times. There are other rows that are present as well, such as client0, Client_Type0, Unknown0, Active0, Obsolete0, Reassigned0, Decommissioned0, and Management0. The non-existant client (at least in the MECM console) has several rows within this system discovery table. I suspect that it's somehow interfering with MECM's ability to push out a valid PXE image. I reviewed the maintenance tasks in MECM, and one of them is "Delete Obsolete Client Discovery Data". Is there a way that this can be run manually instead of on a schedule? OR can I force removal of the client? I suspect our weird behavior is entirely related to the device being removed from the console, but still present in the database. Any help would be welcomed. Thank you!
