Store configuration for Kolide custom ATC tables #1761
Conversation
RebeccaMahany
force-pushed
the
becca/kolide-atc
branch
from
dcc2af1 to
86b7f35
Compare
RebeccaMahany
force-pushed
the
becca/kolide-atc
branch
from
86b7f35 to
5404575
Compare
ee/agent/knapsack/knapsack.go
Outdated
| @@ -87,6 +87,10 @@ func (k *knapsack) AgentFlagsStore() types.KVStore { | |||
| return k.getKVStore(storage.AgentFlagsStore) | |||
| } | |||
|
|
|||
| func (k *knapsack) AtcConfigStore() types.KVStore { | |||
There was a problem hiding this comment.
This will end up in a new bucket, right? Do we want that vs I-don't-know
There was a problem hiding this comment.
Yes, new bucket! I think the other option would be making it an agent flag and it'd live alongside the other agent flags, which I'm not opposed to either. A new bucket and subsystem felt like it gave me a little bit more room to iterate.
There was a problem hiding this comment.
I'm game to try....
In a world where we have 4 katc tables, what will that look like? I imagine they'll be a DSL where we have a table name, and then some parameters defining it.
So if that was in the agent flags, we'd have katc => { table1 => {...}, table2 => {...} } and if it was in a bucket we'll see entries for each table? I guess it's kinda all the same.
There was a problem hiding this comment.
Yeah, the options are exactly as you describe. They both feel ok to me -- let's try this and it shouldn't be painful to switch if we change our minds later 🙂
This is part 1 of enabling ATC for non-vanilla sqlite files (indexeddb, sqlite where the data has been compressed with snappy, dataflatten/exec tables).
This PR gets some boilerplate out of the way:
I don't love "Kolide custom ATC" -- I'm happy to rename if anyone has a better suggestion.