1 |
On 7/15/22 11:46 PM, J. Roeleveld wrote: |
2 |
> Hmm... interesting. I will look into this. |
3 |
|
4 |
:-) |
5 |
|
6 |
> But, it needs the agent to be running, which will make it tricky for |
7 |
> automation. |
8 |
|
9 |
Why can't automation start an agent? Why can't there be an agent |
10 |
running that automation has access to? |
11 |
|
12 |
> (I have some scripts that need to do things on different systems in |
13 |
> a sequence for which this could help) |
14 |
|
15 |
:-) |
16 |
|
17 |
> I know, which is why I was investigating automating it. The passwords |
18 |
> are too long to comfortably copy by hand. |
19 |
|
20 |
I assume that you mean "type" when you say "copy". |
21 |
|
22 |
> I will definitely investigate this. They sound interesting. I'd set |
23 |
> the validity to a lot less if this can be automated easily. |
24 |
|
25 |
Yes, it can be fairly easily automated. |
26 |
|
27 |
One of the other advantages of SSH /certificates/ is when you flip |
28 |
things around and use a /host/ certificate. Clients can recognize that |
29 |
the target host's certificate is signed by the trusted SSH CA and not |
30 |
prompt for the typical Trust On First Use (TOFU) scenario. Thus you can |
31 |
actually leverage the target host SSH fingerprint and not need to ignore |
32 |
that security aspect like so many people do. |
33 |
|
34 |
> Added to my research-list. |
35 |
|
36 |
:-) |
37 |
|
38 |
|
39 |
|
40 |
-- |
41 |
Grant. . . . |
42 |
unix || die |