Random Hosts are added to New group

[fry_p]

Server

• FOG Version: 1.3.0 *OS: Ubuntu 14.01

Client

• Service Version: 0.11.7
• OS: Windows

Description

I should have submitted this before the official 1.3.0 release, so I apologize for the tardiness of this report. Perhaps I am doing something wrong, but whenever I go to search for hosts for a specific lab or room in my district (using the host search tab) and use the select all check box near the top of the list to create a new group, it seems to add some random hosts when I create the group. I can provide logs/screens next week, but my desired group of 25 hosts in a lab turns into a group of 100 hosts, 25 of which are the desired members and 75 seemingly random hosts that were recently registered and imaged.

[Wayne Workman]

@fry_p you might want to clean your database, crap data in it makes some of the code behave bad. Look in the wiki for Troubleshoot Mysql. There’s some commands at the bottom.

[falko]

I too have seen this once or twice, but many revisions ago. And it was only adding the very most recently registered devices (maybe 2-4) to groups. But since then my fog server died and I have rebuilt adding a storage node and haven’t seen this happen in a while since keeping up with the revisions of the development. I actually forgot about it…

I will also check this again, although recently I registered I have registered quite a few new machines and created a few groups and not seen this happen. I’m on Debian 8, FOG 1.3.0

[Wayne Workman]

@fry_p you might want to clean your database, crap data in it makes some of the code behave bad. Look in the wiki for Troubleshoot Mysql. There’s some commands at the bottom.

[falko]

@Wayne-Workman good to know thanks, guessing my rebuild and host import cleaned this up.

[fry_p]

@Wayne-Workman That seems to have done it. Thanks! Is there anything I can do to prevent this crap data, or reduce it?

[Tom Elliott]

@fry_p crap data should already be fixed. If I were a guessing man the crap data is from older builds. There were some cases were in the older code it would create new data for non-existent items. When those new items became the same as the old crap they took on the characteristics you were seeing.

[fry_p]

@Tom-Elliott I didn’t consider that. I had never done maintenance on the SQL stuff prior to this. Thanks for the info.

[Tom Elliott]

I feel I should delve a little more detail into this, for others understanding and ease of understanding.

The “crap data” I’m referring to would be entries for items “in the future” if you.

For example, let’s say you created a snapin and associated that snapin to a group of hosts. There was bad code introduced a while ago that mistakenly created entries in a fashion allowing the potential of hosts, snapins, groups, (whatever it may have been) to be associated to items that did not exist.

Let’s say the snapin you created had a snapin ID of 1. If you have 5 hosts with that snapin and one of the hosts got updated, it would create a new entry. For “static” systems that never had any new hosts added, removed, or whatever this would not have caused any issues. Namely because the host (or whatever item it may have been) did not, and would not have existed. Now, this is an unrealistic scenario and was reported fairly often for that span of code updates. That said, one of the things that would’ve happened would be when a new host was added, the bad (future if you will) already existed. So that new host would suddenly have things associated to them without you having done anything.

Hopefully that all makes sense.

[Wayne Workman]

@falko Probably.