Closed mikolaj-Rianta closed 5 years ago
@mikolaj-Rianta I think I see an issue on our end - we shouldn't be generating potentially "failing" tests for encrypt tests, only for decrypt. I'm not sure if this is the entirety of the issue you experienced or not, but I'll put in a correction in the next release to demo and we can try again.
Russ,
Thanks a lot! Really appreciate it.
Mikolaj
From: Russ Hammett notifications@github.com Sent: Tuesday, April 2, 2019 2:23 PM To: usnistgov/ACVP ACVP@noreply.github.com Cc: mikolaj-Rianta mikolajb@rianta.ca; Mention < mention@noreply.github.com> Subject: Re: [usnistgov/ACVP] AES-XPN test vector expects "testPassed" : false for Encrypt tests!? (#591)
@mikolaj-Rianta https://github.com/mikolaj-Rianta I think I see an issue on our end - we shouldn't be generating potentially "failing" tests for encrypt tests, only for decrypt. I'm not sure if this is the entirety of the issue you experienced or not, but I'll put in a correction in the next release to demo and we can try again.
— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/usnistgov/ACVP/issues/591#issuecomment-479132352, or mute the thread https://github.com/notifications/unsubscribe-auth/Au7XTdZ2nVmRQZJf2_a3_lsmxBm8gHCCks5vc5_pgaJpZM4cYyFG .
This fix has been deployed to demo. Let me know if there are any issues.
Thanks a lot! I'd like to give it a try but there seems to be an issue with JWT when I try POST/GET. I've created a new issue for that. This is the reply I'm getting from the server: "error" : "vsId was not found in the JWT"
Unfortunately I'm still getting "testPassed" : false in 'Encrypt' tests. Today (Apr 4, 2019 @ 10:27 AM) I've created a new testSessionId 2733. Requesting expected results for vectorSet 19750 I'm getting "testPassed" : false for encrypt tcId 1, 2, 3, 9, and 11
Hrm... I'll take a look, thanks!
@mikolaj-Rianta I haven't been able to reproduce this locally, I'm going to guess that the build that went out to demo was just prior to the commit that fixed the issue, where I thought it would have included it.
I'll keep this issue open until our next demo push, at which point we can take another look.
Sorry for the confusion!
This has been deployed to demo. Let us know if there are any issues.
In acvVersion = '1.0', I'm not able to access any of my sessions: e.g. trying to get expected results for testSession 2733 vectorId 19750 returns the following error: "error" : "Cannot properly validate tsId because the claim type isn't supported"
this seems like a separate issue than #607?
Thanks for confirming. I've opened new issue #617
@mikolaj-Rianta just wanted to double check since there have a few updates in here this issue that were due to other issues - can you confirm (with a new session since previously generated JWTs are invalid from prior to the release) that you're no longer seeing failure scenarios on aes-xpn encrypt?
Unfortunately I'm not able to create a new session for AES-XPN here is the error I'm getting: "Incorrectly formatted JSON (18:8): unknown algorithm: AES-XPN with mode: null" It all used to work before rel 1. Please take a look.
Content-Type: application/json
[
{"acvVersion": "1.0"},
{
"isSample" : true,
"algorithms": [{
"algorithm": "AES-XPN",
"revision": "1.0.0",
"direction": ["encrypt", "decrypt"],
"keyLen": [128],
"payloadLen": [1576],
"ivLen": [96],
"ivGen": "internal",
"ivGenMode": "8.2.1",
"aadLen": [200],
"tagLen": [128],
"saltGen": "external",
"saltLen": 96
}]
}
]
Note that the revision should be 1.0
, not 1.0.0
Thanks! It worked
Just requested the expected results for the AES-XPN test vector and I don't see "testPassed" : false for 'Encrypt' tests any longer. Thank you
thanks! 👍
I've submitted my results, testSession 3012 vectorId 20665. Before submission I compared my results with the expected results and the files were identical (except for white spaces) yet when I query the results I get:
"results" : {
"vsId" : 20665,
"disposition" : "incomplete"
},
"retry" : 30
I'll take a look.
A fix has been pushed to demo
Thanks! Should I create a new session or wait for the new release?
can you submit a new capability registration? I don't recall if the currently implemented features allow for failed vector sets to have their answers submitted again (granted this failure was no fault of yours)
Thanks a lot! I'll create a new session 3012 cannot be updated: "error" : "Results for Vector Set id: 20665 cannot be updated with status: KAT_RECEIVED"
Created new session 3027 vectorId 20737, submitted my results and got a pass! Thanks for such a quick turnaround!
no problem, glad it's working!
testSessionId 2131
vsId 19030
Algorithm registration
Endpoint in which the error is experienced https://demo.acvts.nist.gov/acvp/acvp/v1/testSessions/2131/vectorSets/19030/expected
Expected behavior We are not in a position to mark Excrypt tests as failed we can do it for Decrypt tests
Additional context Submitting test results generates the following response:
If you look at artifacts/draft-fussell-acvp-spec-00.html 10.11.4.1 there is no “error” defined