Pending Hosts list is empty

[Tom Elliott]

@x23piracy I don’t know, but if I had to guess i couldn’t give a good status.

[Tom Elliott]

Please run as @Wayne-Workman suggested, followed by the hostPending cleanup.

Then let it run for a bit. I’m suspecting something else is injecting data at “strange” times. The reason why you’re seeing funny mac’s is because at the point that it’s trying to do something there is no network available. Maybe try putting your fogservice startup on a slight delay of 3 minutes?

Just a suggestion based on what I understand of the information.

[x23piracy]

@Tom-Elliott said in Pending Hosts list is empty:

Please run as @Wayne-Workman suggested, followed by the hostPending cleanup.

Then let it run for a bit. I’m suspecting something else is injecting data at “strange” times. The reason why you’re seeing funny mac’s is because at the point that it’s trying to do something there is no network available. Maybe try putting your fogservice startup on a slight delay of 3 minutes?

Just a suggestion based on what I understand of the information.

I have the older fog client 11.4 in my base image (not updated yet) maybe this i making problems. Today i deployed an image and fog service wasn’t there anymore, so i tried to repair fog client and getting cannot install ca cert error, i fully deinstalled and reinstalled from fog website.

Since some RC’s renaming and domain joining isn’t working propperly, i will try to embed latest client in my current image to see if this helps.

[Tom Elliott]

@x23piracy I don’t think it has anything to do with the rc’s or the version of the client any more.

Potentially by cleaning out these bad entries and ensuring your nodes/servers are all updated to the most current, I imagine this issue will no longer be present.

My “theory” on what was causing the bad entries is still just a theory. By trying with a delayed startup of the fog service we should notice relatively quickly if the issue is still present or not. Of course this means updating all your hosts to have the same delayed start, which I don’t think you would want to do right now.

I suppose it’s possible some of the older FOG Clients (prior to 0.11) could cause this issue as well, but i’m going to lean that what you saw was simply older RC versions cause I’m not a very good programmer :(.

[x23piracy]

@Tom-Elliott said in Pending Hosts list is empty:

@x23piracy I don’t think it has anything to do with the rc’s or the version of the client any more.

Potentially by cleaning out these bad entries and ensuring your nodes/servers are all updated to the most current, I imagine this issue will no longer be present.

My “theory” on what was causing the bad entries is still just a theory. By trying with a delayed startup of the fog service we should notice relatively quickly if the issue is still present or not. Of course this means updating all your hosts to have the same delayed start, which I don’t think you would want to do right now.

I suppose it’s possible some of the older FOG Clients (prior to 0.11) could cause this issue as well, but i’m going to lean that what you saw was simply older RC versions cause I’m not a very good programmer :(.

from the past i kept the delayed start, this is still in my image base. this can’t be the reason.

So all my actual clients start the service delayed, anyway i can change the service start behaviour with pdq deploy when ever i want

i cannot exactly tell since when its not working fine anymore but only a few of the last clients cleanly renamed and joined the domain, this was working so perfectly before.

[Wayne Workman]

I have the OP issue now… ugh… I’m going to dig into it in a minute.

[x23piracy]

@Wayne-Workman said in Pending Hosts list is empty:

I have the OP issue now… ugh… I’m going to dig into it in a minute.

what i did today was borrowing a hp x2 port rep to make a pxe boot with a new device without own port rep, therefore i had duplicate mac and pxe boot menu told me host already registered, sure same mac!, so i just renamed the host name of the existing one, deployed the image and after that i deleted the host from fog (to get sure i don’t produce accidents because of the port rep’s mac adress with the original client), little later i had that pending message on home page, sure client was knocking at fog door again, then i detected the error with the empty list after trying to approve it.

[Wayne Workman]

@Tom-Elliott we have three computers that are causing our host pending issue. All three of them were last deployed today.

So, something with the imaging process is causing them to change to pending.

MariaDB [fog]> select hostID, hostName, hostPending from hosts where hostPending = '' or hostPending = '1';
+--------+------------+-------------+
| hostID | hostName   | hostPending |
+--------+------------+-------------+
|   2184 | ET5Q197    |             |
|   2186 | ET5Q198    |             |
|   2185 | L530Win10E | 1           |
+--------+------------+-------------+
3 rows in set (0.00 sec)

MariaDB [fog]> select * from hostMAC where hmPending = '' or hmPending = '1';
+------+----------+-------------------+--------+-----------+-----------+----------------+-----------------+
| hmID | hmHostID | hmMAC             | hmDesc | hmPrimary | hmPending | hmIgnoreClient | hmIgnoreImaging |
+------+----------+-------------------+--------+-----------+-----------+----------------+-----------------+
| 8358 |     2184 | 18:db:f2:0e:25:06 |        | 1         |           |                |                 |
| 8359 |     2185 | 9c:4e:36:5a:11:a1 |        | 1         |           |                |                 |
| 8362 |     2186 | 18:db:f2:13:49:04 |        | 1         |           |                |                 |
+------+----------+-------------------+--------+-----------+-----------+----------------+-----------------+
3 rows in set (0.00 sec)

MariaDB [fog]>

[Tom Elliott]

@Wayne-Workman I only see one pending host

[Wayne Workman]

@Tom-Elliott I set all hosts to ‘0’ earlier. Now there are suddenly 3 that are not set to ‘0’. Nobody deleted them. The only thing they have in common is that they were imaged today.

[Tom Elliott]

@Wayne-Workman that doesn’t matter

[Wayne Workman]

@Tom-Elliott it doesn’t matter that I set the hosts to 0 and they changed on their own to null ?

[Tom Elliott]

@Wayne-Workman correct. Only hosts with a pending code of 1 are evaluated as pending

[Tom Elliott]

@Wayne-Workman And they’re changing “on their own” because every cycle the db get’s updated data. Everytime there’s a checkin, the AD information get’s updated. Because it’s a part of the host, ALL tables of the host get updated, not just the one, it’s how the system works.

Now, if your empty string items are coming back as pending there would be cause for concern, but the checks are based on “loose” checking.

[Tom Elliott]

I set one of my hosts to pending and I see what yo’re seeing.

I don’t know german very well so I didn’t know what was being said there. Now I see it says:

“Requires templates to process”

I fixed it. Will be fixed for RC-21.