GlobalSign: Failure to revoke noncompliant ICA within 7 days
Categories
(CA Program :: CA Certificate Compliance, task)
Tracking
(Not tracked)
People
(Reporter: arvid.vermote, Assigned: arvid.vermote)
Details
(Whiteboard: [ca-compliance] [ca-revocation-delay])
Attachments
(1 file)
127.80 KB,
application/pdf
|
Details |
User Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:68.0) Gecko/20100101 Firefox/68.0
Assignee | ||
Comment 1•4 years ago
|
||
Creating this bug to capture GlobalSign not able to revoke all CA affected by https://bugzilla.mozilla.org/show_bug.cgi?id=1649937 within the 7 days time frame as set forth by section #4.9.1.2 of the SSL Baseline Requirements. We will post the incident report within the week.
Comment 2•4 years ago
|
||
Congrats on being the only CA so far to follow instructions :)
Assignee | ||
Comment 4•4 years ago
|
||
How your CA first became aware of the problem (e.g. via a problem report submitted to your Problem Reporting Mechanism, a discussion in mozilla.dev.security.policy, a Bugzilla bug, or internal self-audit), and the time and date.
We became aware of this problem when we reviewed the issuing CA affected by https://bugzilla.mozilla.org/show_bug.cgi?id=1649937 and internal discussions on remediation plans made it apparent that we would not be able to revoke all affected issuing CA within the timeline as set forth by the Baseline Requirements 4.9.1.2.
A timeline of the actions your CA took in response. A timeline is a date-and-time-stamped sequence of all relevant events. This may include events before the incident was reported, such as when a particular requirement became applicable, or a document changed, or a bug was introduced, or an audit was done.
Time (UTC) | Activity |
---|---|
July 1 2020 21:06 | OCSP EKU security issue disclosed on mozilla.dev.security.policy - refer to https://bugzilla.mozilla.org/show_bug.cgi?id=1649937 |
July 7 2020 12:00 | Second board & leadership debriefing on https://bugzilla.mozilla.org/show_bug.cgi?id=1649937 and discussion on remediation activities for internal CA |
July 7 2020 17:38 | Final conclusion we will not be able to revoke all issuing CA affected by the OCSP EKU security within the timelines set forth by section #4.9.1.2 of the SSL Baseline Requirements and creation of this ticket |
Whether your CA has stopped, or has not yet stopped, certificate issuance or the process giving rise to the problem or incident. A statement that you have stopped will be considered a pledge to the community; a statement that you have not stopped requires an explanation.
As detailed in https://bugzilla.mozilla.org/show_bug.cgi?id=1649937#c9 GlobalSign has ceased including the OCSP signing EKU in any newly generated issuing CA.
In a case involving certificates, a summary of the problematic certificates. For each problem: the number of certificates, and the date the first and last certificates with that problem were issued. In other incidents that do not involve enumerating the affected certificates (e.g. OCSP failures, audit findings, delayed responses, etc.), please provide other similar statistics, aggregates, and a summary for each type of problem identified. This will help us measure the severity of each problem.
Following issuing CA affected by the OCSP EKU security issue were not revoked within the timelines set fort by section #4.9.1.2 of the Baseline Requirements.
ID | CN | SHA256 | crt.sh | Revocation date | Destruction date |
---|---|---|---|---|---|
1 | AbbVie AATL ICA 2020 | 154E4834B28D4FB1F90FEE935D0DDE46C45A177FC1425A028C685C32855A85AD | https://crt.sh/?id=2369948023 | October 21 2020 | October 21 2020 |
2 | CRB Group SMIME CA 2019 | 6A5F4C1678CA65E59F060D57CDFF665065314861D53A8E7D1450CA92D96CA102 | https://crt.sh/?id=2029982659 | October 21 2020 | October 21 2020 |
3 | DexKo Global SMIME CA 2019 | ABC86706C98D6BF67372F908EC01ADF631B191D733AE89F8343EB047B108144B | https://crt.sh/?id=2029984306 | November 18 2020 | November 18 2020 |
4 | GlobalSign AATL Partners CA 2019 | 83FC891B350D9E0D7EBE6DD2A6BFE3D0B0F4653FCA048615A5DEEBBC039A3F66 | https://crt.sh/?id=1436918881 | October 21 2020 | October 21 2020 |
5 | GlobalSign Issuing CA for AATL Partners 2019 | 67C46DC17762667844F1596089375FF45E05C2B316C89499F6E7FAB78C8F0379 | https://crt.sh/?id=1703475173 | September 16 2020 | October 21 2020 |
6 | GlobalSign Qualified Time Stamping CA 2019 | 74ABE5E5CCEB75491FF72C4CF325405D8ADBFE390E189CF430BA60E62798878E | https://crt.sh/?id=1490728721 | N/A | July 28 2020 |
7 | Qu\C3\A1litas Compa\C3\B1\C3\ADa de Seguros S.A. de C.V. | B716B089FE4E53D1A2EF7BA57AC85E68EC722CF61052C25A59626AD3B15C5F40 | https://crt.sh/?id=1814826066 | October 21 2020 | October 21 2020 |
8 | Ford Motor Company - Enterprise Issuing CA01 | 4C241CFE3D3FFB60CA88D6B06A552AB1CF0EF7D8D2E08DA15282B55192EBBD29 | https://crt.sh/?id=392882654 | Parent revoked on July 11 2020 | N/A |
9 | Ford Motor Company - Enterprise Issuing CA01 | 3802E424516F78EEAC329AAE9B1F60A412DBE1D5B095D7AC9DC0DCDDE3C1F5FB | https://crt.sh/?id=306624237 | Parent revoked on July 11 2020 | N/A |
10 | GlobalSign CodeSigning CA - G3 | 4047C9D69260C07213BCB8608A7EC5E2838A56B79F67847812EAC0778D0D27F1 | https://crt.sh/?id=157564305 | April 21 2021 | April 21 2021 |
11 | GlobalSign PersonalSign 1 CA - G3 | F068DEAA18CC02D5A8BE35CB8338327910291F6E62E7216A934764A1ABA4A800 | https://crt.sh/?id=2369948051 | July 28 2020 | July 28 2020 |
12 | GlobalSign PersonalSign 2 CA - G3 | 925EE7D5A22AD7FBE9BAB54D7C8D0B9A74F7E35A8AF6AF645E2E8C3519A7092F | https://crt.sh/?id=2369947954 | July 28 2020 | July 28 2020 |
13 | GlobalSign PersonalSign 2 CA - SHA256 - G3 | B778748A792B8F91F04B01BAFC31A31ED7EF6A712AFF80B6610D9AADEE207ADF | https://crt.sh/?id=24592899 | February 24 2021 | February 24 2021 |
14 | GlobalSign PersonalSign 2 CA - SHA256 - G4 | 27D6FDAF80297846DFEFF82E7F58B9A48AC9E3EE93A112B1BBE243EE1A97447C | https://crt.sh/?id=2839140428 | July 11 2020 | July 28 2020 |
15 | GlobalSign PersonalSign 2 ECC CA SHA 384 - G4 | 46038F6326228CDB56619C52266613DA04C8CA499E0D03B0EDCFFC110D5CFC70 | https://crt.sh/?id=405618313 | July 28 2020 | July 28 2020 |
16 | GlobalSign PersonalSign 2 RSA CA SHA 384 - G4 | 5CDD809CF44F5F8665EAC15055504C5B06B787AC18294505BDBAB4A77E50D776 | https://crt.sh/?id=405618295 | July 28 2020 | July 28 2020 |
17 | GlobalSign PersonalSign 3 CA - G3 | B1FE3AEBF963A7880E74B0B0556681EA8B1CCCE3E69A7D3B10A68ACBE86E48A1 | https://crt.sh/?id=2369948436 | July 28 2020 | July 28 2020 |
18 | GlobalSign SMIME CA 2018 | C8192C32F7B49C7F32A1CA001595A7F9E36C9E72058D6EAA1BAB7752A8C16718 | https://crt.sh/?id=549505576 | September 16 2020 | October 21 2020 |
19 | JCAN Public CA1 - G4 | 7B464DC384FDB1A525C2CC279ED0C7CFAD24BECF72C46A7D7093D157C217607E | https://crt.sh/?id=163676419 | Parent revoked on July 11 2020 | N/A |
20 | NAESB Issuing CA - SHA384 - G3 | 128DED1A8AD60C24B4254E31DB94FC4392BF93ED5434472AA43A0B9856106068 | https://crt.sh/?id=2369948019 | July 28 2020 | October 21 2020 |
21 | NAESB Issuing CA - SHA384 - G3 | 0986B5A1C7314EFB04FB648B9E2B57CF4842FD1D4345D28E52094C90A9FECBFE | https://crt.sh/?id=18068129 | September 30 2020 | October 21 2020 |
22 | SHECA DV Secure Server CA | 393B8B15CABC3886FB2E416495D63C8BADD8DCAF87552076C8A0A9637C24DE47 | https://crt.sh/?id=1225556701 | Parent revoked on July 11 2020 | N/A |
23 | SHECA EV Secure Server CA | 147C447FEEB86202B503314FCAF0036BEAAEF437C39B56B358EC446A9D20387F | https://crt.sh/?id=1229139434 | Parent revoked on July 11 2020 | N/A |
24 | SHECA OV Secure Server CA | 77EAC476453CB732257FF166A5EBD1656CB1F673B68E28DF41774133979FA2A4 | https://crt.sh/?id=1225556702 | Parent revoked on July 11 2020 | N/A |
25 | ATT Organization Validated CA 2019 | 7AA45D6F5B14DAB1C6844C19C2804E14B5811E6EDE1F02B0AEF065A7B359C68F | https://crt.sh/?id=1490728430 | September 30 2020 | October 21 2020 |
26 | CrowdStrike OV SSL Issuing CA 2020 | AE03B9AD17106A28785830B1DCD636797C4C64D81CB8D161595DBAF83433E64C | https://crt.sh/?id=2839140453 | Parent revoked on July 11 2020 | N/A |
27 | DPDHL Global TLS CA - I4 | 94C663E9EA5C27EE4F64127F9B425863E991A9E156C07DF1A00803AE31764162 | https://crt.sh/?id=1814823951 | September 30 2020 | October 21 2020 |
28 | DPDHL User CA I3 | AF1898D7F0638751C075D0142D4E2A0EA731FC622324F153FE1BF3B6AFD9AF13 | https://crt.sh/?id=1596016275 | Parent revoked on July 11 2020 | N/A |
29 | DPDHL User CA I3 | 2E0191751CA0CBA81C3A6338DEE1A02B8D6BCC4F1F8261B809BCCE7ABAF1A43D | https://crt.sh/?id=12729527 | Parent revoked on July 11 2020 | N/A |
30 | DPDHL User CA I3 | BCE3A5BD8D9082636C5BFE3E0B71ACEE551E24E3BD035887D2661ADA65AFF484 | Not in crt.sh | Parent revoked on July 11 2020 | N/A |
31 | DPDHL User CA I3 | F037621405E0F356507E239FADD647842D3B50857C3CFF840859174F72F6FD18 | https://crt.sh/?id=329514052 | Parent revoked on July 11 2020 | N/A |
32 | DPDHL User CA I4 | C25C4EDBC36E3FB7C3D937BEE9F2D29E36AFB07CFA3188262E0D5FDC919E0D77 | https://crt.sh/?id=2369948075 | November 18 2020 | November 18 2020 |
33 | Giesecke and Devrient CA | 632FD697BACAF1ED232517EC9B7622B7C25E1448B0CC626B33286719E351CE8A | https://crt.sh/?id=196919504 | Parent revoked on July 11 2020 | N/A |
34 | GlobalSign CA 4 for AATL | EBA34C7B109671614C367E1DE075124C3954CE19F85FACF61090EC319F7F1A7F | https://crt.sh/?id=1229139435 | December 31 2020 | January 20 2021 |
36 | GlobalSign CA 5 for AATL | 306E9739E3458FF4546877B704B2E3905E58B235D64E32F4F026AC91B7295D15 | https://crt.sh/?id=408789250 | December 31 2020 | January 20 2021 |
37 | GlobalSign CA 6 for AATL | BE1FFC0E1FF6088104F43E327E7C7DC72A9CA7B0DF05793123ABE32DEACEE76F | https://crt.sh/?id=2369988390 | December 31 2020 | January 20 2021 |
38 | GlobalSign CA for AATL - SHA256 - G3 | 2E8820DC0EAFAE3D6D285C057ECE14470B377438B002CEDD4C72B4F343A54F43 | https://crt.sh/?id=2369947889 | N/A | January 20 2021 |
39 | GlobalSign CodeSigning CA - SHA256 - G2 | BE40813869AB27A071D12AD6A8830583EBC3B618E3F2346359F4B11A1C9434EE | https://crt.sh/?id=1703475054 | October 21 2020 | October 21 2020 |
40 | GlobalSign CodeSigning CA - SHA256 - G3 | FB54EEA9BCE8E9EA9782154F3D414277FB709F49B947D73978AC278546C2CE03 | https://crt.sh/?id=26749929 | April 21 2021 | April 21 2021 |
42 | GlobalSign Extended Validation CodeSigning CA - SHA256 - G2 | 1E864278C20881B671C0C6D2E14B61150AD1F13CF92C6EC14B550DCBC47E1541 | https://crt.sh/?id=1703475088 | October 21 2020 | October 21 2020 |
43 | GlobalSign Extended Validation CodeSigning CA - SHA256 - G3 | DD038E87E0B4D2C369680D3DE78638AB39FC1D7E50632996921101768DB8D4D8 | https://crt.sh/?id=41285443 | April 21 2021 | April 21 2021 |
45 | GlobalSign HV RSA DV SSL CA 2018 | 54C37A8E853FD1D6378D378B939307EC321A31CC1A5A89E7180633BC13F18762 | https://crt.sh/?id=970082980 | October 21 2020 | October 21 2020 |
46 | GlobalSign PersonalSign 1 CA - SHA256 - G3 | F5D2D2BA6817A7A9AA0E21354BBF0E6F95C5E287EE88CF2F279F0FFEC4EDAC15 | https://crt.sh/?id=147619379 | February 24 2021 | February 24 2021 |
47 | GlobalSign PersonalSign 3 CA - SHA256 - G3 | 701B432AC0CDD4D9CF95B4B884C32BF5CCA90D44E0161ABD13B934D68E380472 | https://crt.sh/?id=163079175 | February 24 2021 | February 24 2021 |
48 | GlobalSign PersonalSign Partners CA - SHA256 - G2 | 4E707867946AC05343C6BA8FF121EA66A758037913257A8EE4974350D39A1034 | https://crt.sh/?id=2369948041 | January 20 2021 | January 20 2021 |
50 | GlobalSign PersonalSign Partners CA - SHA256 - G2 | 118262C2088EE1528E20D836D2070854707C0D8F8E80FBE396F9ECD4B9141B5B | https://crt.sh/?id=12715740 | September 30 2020 | January 20 2021 |
51 | GlobalSign Qualified CA 1 | F5709A2D2F68B53BF6F645BB178ADF95346F89FDA5C63BFDE08042A26492AAB2 | https://crt.sh/?id=509714293 | February 24 2021 | February 24 2021 |
57 | GlobalSign RSA DV SSL CA 2018 | 9E898ED03FA46969690DAD73C7296675045FF9B5A0100A399BEB8435A98F5185 | https://crt.sh/?id=970083106 | January 20 2021 | January 20 2021 |
58 | GlobalSign RSA EV QWAC CA 2019 | EDC734C501501DC7A27448FA02C74931F8578BF297B173F34B841E82C6691926 | https://crt.sh/?id=1490728500 | February 24 2021 | February 24 2021 |
60 | GlobalSign Timestamping CA - G3 | 95C6A747DD0BC755A1941827E894B8083592241B792541E2EB1B30FB9B13F57F | https://crt.sh/?id=2392141070 | N/A | May 28 2020 |
61 | GlobalSign Timestamping CA - SHA256 - G3 | BE33D1C57EBDDD927B57BDB604BE457B552FE568E7F3DCBA093C39ED1C30A239 | https://crt.sh/?id=2369948437 | N/A | May 28 2020 |
62 | LinQuest SMIME CA 2020 | 113138DD7B216725840238E2D7EEECB3738DB139064B24CB853FC270A49E6057 | https://crt.sh/?id=2369948433 | November 18 2020 | November 18 2020 |
63 | NAESB Issuing CA - SHA384 - G4 | C4C7C436BD88E8E68DB00297DF83ACC819E198639BA00522C8E3245876898523 | https://crt.sh/?id=2369948432 | February 24 2021 | February 24 2021 |
64 | RNP ICPEdu OV SSL CA 2019 | 42CFDDA6F660B8E5B4C1C411965A4519312559E3262F8DB69D2DAE17B26B3BA3 | https://crt.sh/?id=1476651440 | Parent revoked on July 11 2020 | N/A |
65 | Trafigura PTE Ltd S/MIME ICA 2020 | 5E7FCB9C97BDA56993B1658D120232761D665A3644534300FA6A5BEC5E0D5795 | https://crt.sh/?id=2369948428 | July 11 2020 | July 28 2020 |
66 | GlobalSign CodeSigning CA - G2 | FFFE077503FD72F0E5338B0A7B4E218E7D1FF82E493E7E852AE51AA1C7585D17 | https://crt.sh/?id=1476651569 | October 21 2020 | October 21 2020 |
67 | DPDHL User CA I3 | EBE87BB4188502709F444055259ABB22BC51B88C908419A13559DFC8EF6630D1 | https://crt.sh/?id=12729526 | Parent revoked on July 11 2020 | N/A |
68 | Ford Motor Company - Enterprise Issuing CA01 | CF73B52D041B7309B439D16247414B90C9D26E44E38748A36500D5829B5187F9 | https://crt.sh/?id=215376217 | Parent revoked on July 11 2020 | N/A |
69 | Ford Motor Company - Enterprise Issuing CA01 | 3B9668F59F55FA3838FC2A3B80B7F9B5B13D1A46F1EAA6E0BCFF04C54198056C | https://crt.sh/?id=215376215 | Parent revoked on July 11 2020 | N/A |
75 | Liberty University External Issuing CA 01 | 1F91212C6BFC333C6EB52A685525E1E5B9E3AC1EF7A5A86649F5F95C721D8898 | Not in crt.sh | Parent revoked on July 11 2020 | N/A |
76 | Liberty University External Issuing CA 01 | CA005AA75E33594BD1DEDC584E1E74E5198EBB1DE88929ED4F3E2E9FFCE3873B | https://crt.sh/?id=36391364 | Parent revoked on July 11 2020 | N/A |
79 | Crown Prince Court CA | BF5EDFBEEB85999C5169CBF3F4DB63B679AD2E1E2272FC3795F9F9921E6D0487 | https://crt.sh/?id=7890405 | Parent revoked on July 11 2020 | N/A |
80 | Crown Prince Court CA | A0133BE5B14E02310A2D4BEAB601094F1194EE8BD6FD29DDFE7B9347467C2EEC | https://crt.sh/?id=10105729 | Parent revoked on July 11 2020 | N/A |
81 | Crown Prince Court CA | F164AD5E4CE9EFC0A144CA902EA2ED46C464D2D508CA919A23095CDF30D4DC68 | Not in crt.sh | Parent revoked on July 11 2020 | N/A |
82 | Crown Prince Court CA for AATL | DF45EEAED905C58D730EC5497B59B3AB4CCE7C6459953DF9CA5C1F031AC06DD8 | Not in crt.sh | Parent revoked on July 11 2020 | N/A |
83 | Crown Prince Court CA for AATL | D21076207F79A9B04137D40A4FFE6DD08921CCF49E6EB60277FF4593E076D538 | Not in crt.sh | Parent revoked on July 11 2020 | N/A |
84 | Crown Prince Court CA for AATL | 70BDB19C31F5EF105B29376E35EA5ED8EEBE13CB5C0758C32DFC4C5F7230A173 | Not in crt.sh | Parent revoked on July 11 2020 | N/A |
85 | DPDHL TLS CT CA I3 | 9153E4420DDC7EB4E6E864AA0377DADF4082ECD35052113638E05D3C296BC006 | https://crt.sh/?id=786990298 | Parent revoked on July 11 2020 | N/A |
86 | DPDHL TLS SHA 2 CA I3 | 25BACC40A5392B82AADEA04903905A467121F28220E6F2F7E0FE982AAFC14FA6 | https://crt.sh/?id=8527797 | Parent revoked on July 11 2020 | N/A |
87 | DPDHL TLS SHA 2 CA I3 | 5A405535C112A0A81AF0D2ACCA3C3F9BC1A677586CDBC633CB4F5F778E1A3550 | https://crt.sh/?id=329514048 | Parent revoked on July 11 2020 | N/A |
88 | DPDHL TLS SHA 2 CA I3 | 23A74704D77A03CFD3FF19E62C500848214E6C60FD2AAEF7DCE7A8F9EE9F9232 | https://crt.sh/?id=27823827 | Parent revoked on July 11 2020 | N/A |
89 | DPDHL TLS SHA 2 CA I3 | 1C942A22A016A1E5559DAE77EC5CE8671F98AE0BA4AC2DC259418E8E1E9F94AD | https://crt.sh/?id=9881176 | Parent revoked on July 11 2020 | N/A |
90 | Southern Company External Issuing CA 1 | FB953C4FC0045846D02491C8ECCF387BA34347C17ABB0EA6D59F6DE4D2F1EA04 | https://crt.sh/?id=11501550 | Parent revoked on July 11 2020 | N/A |
91 | Trusted Root CA G2 | 6E32A35B599E9087BB1AB35CE73022EC2E26AF34BE388919419C95700CD8E7FB | https://crt.sh/?id=1862521 | July 11 2020 | N/A |
92 | Trusted Root CA SHA256 G2 | 01FD73EF5E70F526FC9C11F65FE2EE6F7125B3693949227FFD8E459E583C458A | https://crt.sh/?id=3179271 | July 11 2020 | N/A |
93 | CBMM SMIME CA 2019 | C346D9137E05254C6EEAC99AC2F6748A0C5D3AFC6B7B9B1E00C40ADF4D85655D | https://crt.sh/?id=1596016282 | November 18 2020 | November 18 2020 |
94 | Accenture Federal Services External CA | A9C8E971259A2ED6E65F721E07AA967C72C7CDB47C7BE1288D87BF08D2F3580D | https://crt.sh/?id=215376216 | December 31 2020 | January 20 2021 |
95 | EY LLP SHA256 CA - G2 | 1557F65BA61C958B74EFA4A582BBAEBDD62A6D9B65FE95A80D5ED518F46ED87F | https://crt.sh/?id=970084237 | December 31 2020 | January 20 2021 |
96 | GROB-WERKE GmbH und Co. KG SMIME CA | 1608BF87414CDCFAB4279102A19702D9D5996A91329E3DF2F80495473AAD86C6 | https://crt.sh/?id=872293772 | October 21 2020 | October 21 2020 |
97 | Hyperion SMIME CA 2018 | DCD77E34B46D530AEF645A513389CD4FFC0F7D196A115B8F62A5FD0D557D46C6 | https://crt.sh/?id=721305503 | November 18 2020 | November 18 2020 |
98 | Spirit AeroSystems SHA256 CA - G2 | 8609BDCEF95E4A4D426497B5CD8ED4B001C953A5C14471CAAF58FB650DF8ABF0 | https://crt.sh/?id=215376218 | November 18 2020 | November 18 2020 |
99 | HERTZ SHA256 CA - G2 | 64FD7F66B805EF0FAE09DAC06EAD1A9AB5C28E6F24AD0759996D349987FEE7E2 | https://crt.sh/?id=408789253 | October 21 2020 | October 21 2020 |
In a case involving certificates, the complete certificate data for the problematic certificates. The recommended way to provide this is to ensure each certificate is logged to CT and then list the fingerprints or crt.sh IDs, either in the report or as an attached spreadsheet, with one list per distinct problem. In other cases not involving a review of affected certificates, please provide other similar, relevant specifics, if any.Explanation about how and why the mistakes were made or bugs introduced, and how they avoided detection until now.
Refer to the table above.
List of steps your CA is taking to resolve the situation and ensure that such situation or incident will not be repeated in the future, accompanied with a binding timeline of when your CA expects to accomplish each of these remediation steps.
1. Moving non-TLS use cases away from TLS roots
A significant portion of the affected CAs relate to non-TLS use cases: document signing, time stamping, S/MIME, authentication, code signing and esoteric PKI ecosystems such as NAESB, eIDAS and AATL.
As detailed in https://bugzilla.mozilla.org/show_bug.cgi?id=1599788#c4 GlobalSign created a new set of roots March 2020 with the aim to separates different PKI certificate use cases under different roots - the naming scheme (R|E)[0-9]{2} is R|E for the algorithm, [0-9]{2} for the year of root expiry.:
- GlobalSign Secure Mail Root R45
- GlobalSign Secure Mail Root E45
- GlobalSign Code Signing Root R45
- GlobalSign Code Signing Root E45
- GlobalSign Document Signing Root R45
- GlobalSign Document Signing Root E45
- GlobalSign IoT Root R60
- GlobalSign IoT Root E60
- GlobalSign Timestamping Root R45
- GlobalSign Client Authentication Root R45
- GlobalSign Client Authentication Root E45
- GlobalSign Root R46 (TLS)
- GlobalSign Root E46 (TLS)
The above roots were included in our latest WebTrust audits dated June 29 2020. Since the delivery of the WebTrust reports and inclusion in our CPS we have started enrolling the above roots in the relevant root programs. The initial plan was, once sufficient trust and ubiquity within the context of the use case was obtained, to generate future issuing CA and leaf certificates under these new roots as of 2021 with as endgame having fully seperated hierarchies, in order to reduce the scope, WebPKI impact and remediation effort of any future similar compliance incidents.
In the context of remediating the EKU OCSP incident we are accelerating this effort: where obtaining the necessary ubiquity for their use case takes time, we are cross-signing the roots with current generation roots and for those use cases where issuing CA enrollment into an esoteric PKI ecosystem is required (e.g. eIDAS, NAESB and AATL) we have started those activities. Affected issuing CA with a non-TLS use case will be migrated off our current generation of mixed use case roots to dedicated roots, after which we will revoke the affected issuing CA. As of now, future issuing CA with a non-TLS use case will be generated under the next generation of roots (under the condition this does not cause any trust or ubiquity issues).
2. Key management bandwidth
In the context of business continuity and responding to CA key compromise we have long maintained the capability to perform revocation key ceremonies from three different key management locations within a short amount of time (24-72 hours depending on the location). At the time of the OCSP EKU incident only one of these locations had the capability and resources for generating new issuing CA keys and certificates with no lead time (we have a secondary location with that capability but the lead time to perform the actions there is a few days). Given the amount of issuing CA affected and corresponding key management tasks in terms of re-generation, revocation and destruction it became apparent we need to expand our key management bandwidth in order to perform these different key management actions more timely in the future.
We have commenced the activities to render a second key management facility and associated resources capable of performing expand key management tasks with no lead time and ensuring more available bandwidth for future incident response activities. This is expected to be completed November 16 2020.
3. Increased rotation of TLS issuing CA
For our new platform "Atlas" we are currently architecting and building more frequent rotation of TLS ICA so the amount of leafs that chains to them is contained, which will allow a more timely re-issuance of leaf certificates and revocation of their affected issuing CA.
In 2021 we will rotate the GlobalSign TLS issuing CA on the Atlas platform every 6 months and customer-dedicated TLS issuing CA on the Atlas platform every year, further increasing the frequency of GlobalSign TLS issuing CA rotation to every quarter in 2022. Our legacy platforms are expected to be fully deprecated for TLS in 2023.
4. Automation
GlobalSign has been actively promoting APIs for customers to manage their certificates and we are continuing to build on those to expand automation and provisioning.
- We're in the final development of our ACME services which will enable fully automated domain validation, issuance and provisioning of DV and OV TLS certificates for service providers and Enterprise customers. This will be available to select customers in Q3 of this year and to a wider audience in 2021 as more customers use our Atlas platform.
- Our Automated Enrollment Gateway, AEG, is an on-premise proxy solution which automatically provisions TLS certificates based on Active Directory policies using auto-enrollment. AEG also supports non-domain joined platforms via the use of ACME for certificate issuance. AEG permits enterprise customers to centrally manage their endpoints and to initiate and push out new certificates without needing to initiate the issuance or replacement from the web server.
Comment 5•4 years ago
|
||
Some of these CAs stand out as quite long, e.g. "GlobalSign CodeSigning CA - G3", "GlobalSign PersonalSign 2 CA - SHA256 - G3", "NAESB Issuing CA - SHA384 - G4", or even "EY LLP SHA256 CA - G2"
What controls are in place to mitigate risk, and what's GlobalSign doing to provide assurance in the interim of the correctness of those controls and their assessment, e.g. such as the provision of a Detailed Controls Report?
Time (UTC) Activity July 1 2020 21:06 OCSP EKU security issue disclosed on mozilla.dev.security.policy - refer to https://bugzilla.mozilla.org/show_bug.cgi?id=1649937 July 7 2020 12:00 Second board & leadership debriefing on https://bugzilla.mozilla.org/show_bug.cgi?id=1649937 and discussion on remediation activities for internal CA July 7 2020 17:38 Final conclusion we will not be able to revoke all issuing CA affected by the OCSP EKU security within the timelines set forth by section #4.9.1.2 of the SSL Baseline Requirements and creation of this ticket
Could you update your timeline to include all relevant events leading to the delayed revocation? The current published timeline raises the concern that GlobalSign, even with the mentioned remediations, will not be able to revoke their intermediate CAs in time because the first related activity after the problem report is a debriefing at the day before the BR deadline.
Assignee | ||
Comment 7•4 years ago
|
||
(In reply to Matthias from comment #6)
Time (UTC) Activity July 1 2020 21:06 OCSP EKU security issue disclosed on mozilla.dev.security.policy - refer to https://bugzilla.mozilla.org/show_bug.cgi?id=1649937 July 7 2020 12:00 Second board & leadership debriefing on https://bugzilla.mozilla.org/show_bug.cgi?id=1649937 and discussion on remediation activities for internal CA July 7 2020 17:38 Final conclusion we will not be able to revoke all issuing CA affected by the OCSP EKU security within the timelines set forth by section #4.9.1.2 of the SSL Baseline Requirements and creation of this ticket Could you update your timeline to include all relevant events leading to the delayed revocation? The current published timeline raises the concern that GlobalSign, even with the mentioned remediations, will not be able to revoke their intermediate CAs in time because the first related activity after the problem report is a debriefing at the day before the BR deadline.
The supporting overview of events is available in https://bugzilla.mozilla.org/show_bug.cgi?id=1649937#c9.
Time (UTC) | Activity |
---|---|
July 1 2020 21:06 | Security issue disclosed on mozilla.dev.security.policy |
July 1 2020 21:42 | CISO is notified of the post and the fact GlobalSign is affected |
July 1 2020 22:22 | CISO notifies the leadership and compliance team and mobilizes an investigation team |
July 2 2020 19:31 | Investigation and impact analysis completed, initial remediation plan finalized: focus on remediation of hierarchies containing third-party operated issuing CA keys first and remediate the affected issuing CA / keys under GlobalSign control as a second priority |
July 3 2020 01:00 | Key manager and compliance team start preparing for initial re-issuance of issuing CA within hierarchies that contain third-party operated issuing CA keys outside of GlobalSign's controls (Trusted Root) |
July 3 2020 01:00 | Incident team starts preparation activities to work with customers affected by the initial revocation activities |
July 3 2020 13:00 | GlobalSign board & leadership are debriefed by CISO and approve the plan of action: revoke the affected Trusted Root hierarchies (the GlobalSign hierarchies that contain third-party operated issuing CA keys outside of GlobalSign's controls) on July 8 16:00 UTC |
July 3 2020 14:00 | Incident team starts to work with customers affected by the initial revocation activities to re-issue of certificates under alternate hierarchies or prepare for swapping their issuing CA with a new one to be generated on July 5 2020 |
July 5 2020 09:00 | Key ceremony to generate new issuing CA for affected customers under Trusted Root hierarchies |
July 5 2020 11:00 | Key ceremony concluded, start setting up new issuing CA and re-issuance activities for customers |
July 7 2020 12:00 | Second board & leadership debriefing and discussion on remediation activities for internal CA |
July 7 2020 13:00 | GlobalSign receives official letter from a regional infrastructure provider in Japan, which has leafs under the "JCAN Public CA1 - G4" (https://crt.sh/?caid=51044), chained to the "Trusted Root CA SHA256 G2" (https://crt.sh/?caid=1423). The letter detailed that a revocation of the hierarchy on July 8 2020 UTC would have significant impact on the companies' ongoing relief and recovery efforts related to restoring their services in the context of the ongoing flooding disaster on southern part of Japan. |
Assignee | ||
Comment 8•4 years ago
|
||
(In reply to Ryan Sleevi from comment #5)
Some of these CAs stand out as quite long, e.g. "GlobalSign CodeSigning CA - G3", "GlobalSign PersonalSign 2 CA - SHA256 - G3", "NAESB Issuing CA - SHA384 - G4", or even "EY LLP SHA256 CA - G2"
What controls are in place to mitigate risk, and what's GlobalSign doing to provide assurance in the interim of the correctness of those controls and their assessment, e.g. such as the provision of a Detailed Controls Report?
We are currently discussing with a qualified auditor on the options to provide an ISAE3000 SOC2 Type I/II report on the non-performance of OCSP signing by and protections in place for the affected issuing CA & keys. We'll share an update once we have made progress on the matter.
Comment 9•4 years ago
|
||
Checking in for an update/progress on this point? I'm not sure a Type I report would be that useful, but perhaps you can share more details?
It seems a WebTrust Detailed Controls Report would be more relevant?
Assignee | ||
Comment 10•4 years ago
|
||
We thought of reporting on a customized set of controls including relevant controls from existing requirement sets (e.g. WebTrust Criterion 6.8) but also tailored controls not directly included in any of the requirement sets but relevant to the context of this incident and associated risk, to make sure the following information (and other system design features) as reported in https://bugzilla.mozilla.org/show_bug.cgi?id=1649937 are sufficiently covered - please let us know if this is not the expectation:
Knowing that this does not provide any assurance, here are some additional insights into our environment:
- Even if they were intended to be potentially used for that purpose in the long term (OCSP response pre-generation directly from the CA), none of the affected issuing CA and associated keys have effectively been used to sign OCSP responses, we use dedicated OCSP responder systems, certificates, keys and HSMs for that purpose.
- Our CA issuance systems are logically set up in such a way that only outbound push/pull queuing from the CA issuance segment to another network segment is possible. No other interactions between the CA issuance segment and other networks are possible.
- CA issuance systems are stripped of any unnecessary functionality and do not contain the necessary software to generate OCSP responses (short of manual generation using pkcs11-tool or openssl, which would be captured in the audit logs)
- Apart from logical segmentation our CA issuance systems and HSMs are physically in a separate zone (room) with multi-person access requirements, preventing the CA keys to be exposed or exported to other systems such as OCSP responders.
- We maintain configuration records and system change logs for all systems involved in CA operations
- Each key signing action on CA issuance systems is logged
We would also prefer to share the draft system design description and list of controls with the community first prior to starting the audit activities to make the reported matter will match the expectations of relying parties. Given current progress we expect a first draft to be ready by August 21.
I also thought that providing a Type I (to provide assurance on the design of the system) as soon as possible, and a Type II (covering the period from the security risk having been raised to the issue being remediated) would provide additional assurance compared to accepting our claim that we do not use ICA keys for OCSP signing, and only receiving an assurance report after a considerable amount of time (i.e. when the issue is remediated).
Assignee | ||
Comment 11•4 years ago
|
||
Update - we are still working on the system design description and list of controls. In the meanwhile any suggestions, expecations or feedback on the reporting approach are appreciated.
Comment 12•4 years ago
|
||
Arvid: Do you have an update here? Comment #10 mentioned August 21, and while Comment #11 noted a delay, I haven't heard an update.
Assignee | ||
Comment 13•4 years ago
|
||
Please find attached a first proposal of the controls and testing activities to be covered in the custom SOC2 report on the OCSP EKU incident. The idea is to obtain a SOC2 Type II with a period-under-audit covering April 22 2020 through April 21 2021 (the date on which the last keys associated with CA that have the OCSP EKU included will be destroyed).
We have selected the controls which we deem most applicable in the context of the incident and associated risk or where a more specific population (i.e. a random sample or full population of OCSP EKU-containing CA) compared to the regular WebTrust audit makes sense. We have already engaged with a qualified auditor for the execution of this work.
Any feedback on the proposed controls and tests or suggestions for control expansion to provide more assurance are welcomed. We are also in an ongoing discussion with the auditor whether we can also have a SOC3 version of the same report to share with any relying party.
Comment 14•4 years ago
|
||
Arvid: Thanks for attaching this, and apologies for the delay here.
With respect to Control 2.1 of WebTrust for CAs, the wording as part of the test includes the following language:
based on risk assessment and in accordance with the CP and CPS.
This seems to be at odds with the Baseline Requirements, Section 6.2.7, which does not include a risk assessment with respect to protection of CA key material, as it uncategorically expects them to be maintained in such devices. This seems like it can be corrected with regards to the test to be applied.
Control 3.3 is perhaps the most important and relevant to the community here, and I'm somewhat concerned that the test to be applied doesn't describe the set of tests the auditor(s) will be performing in order to reach the necessary assurance level. We'e seen from past incidents some confusion and misunderstanding, and while I'm thrilled to see a GlobalSign internal control explicitly included, I'm also concerned that without understanding the procedures, it may provide limited assurance to relying parties. For example, one could imagine an auditor (incorrectly) examining certificate and CRL issuance databases, and only seeing certificates and CRLs, rather than examining the HSM activity logs and correlating those to said databases and manual ceremonies to ensure no unattributed use of the key material. I think this similarly touches on the tests for Control 7.1
However, I do want to commend GlobalSign for again, by leaps and bounds, providing an exemplary demonstration of the transparency and detail that we expect of all CAs. Comments like Comment #13 are the model for what CAs should be achieving here.
Assignee | ||
Comment 15•4 years ago
|
||
We are currently finalizing the SOC2 controls incorporating the feedback provided in Comment #14, and will share the final version as soon as we've reached consensus with our auditor. The following remedial action related to key management bandwidth has been completed on November 9 2020.
2. Key management bandwidth
In the context of business continuity and responding to CA key compromise we have long maintained the capability to perform revocation key ceremonies from three different key management locations within a short amount of time (24-72 hours depending on the location). At the time of the OCSP EKU incident only one of these locations had the capability and resources for generating new issuing CA keys and certificates with no lead time (we have a secondary location with that capability but the lead time to perform the actions there is a few days). Given the amount of issuing CA affected and corresponding key management tasks in terms of re-generation, revocation and destruction it became apparent we need to expand our key management bandwidth in order to perform these different key management actions more timely in the future.
We have commenced the activities to render a second key management facility and associated resources capable of performing expand key management tasks with no lead time and ensuring more available bandwidth for future incident response activities. This is expected to be completed November 16 2020.
Comment 16•3 years ago
|
||
Please provide a status update on these ICAs. Thanks.
Assignee | ||
Comment 17•3 years ago
|
||
Hi Ben - please refer to https://bugzilla.mozilla.org/show_bug.cgi?id=1649937#c30, the next batch of CAs will be revoked on January 20 2021.
As for following remedial action:
3. Increased rotation of TLS issuing CA
For our new platform "Atlas" we are currently architecting and building more frequent rotation of TLS ICA so the amount of leafs that chains to them is contained, which will allow a more timely re-issuance of leaf certificates and revocation of their affected issuing CA.
In 2021 we will rotate the GlobalSign TLS issuing CA on the Atlas platform every 6 months and customer-dedicated TLS issuing CA on the Atlas platform every year, further increasing the frequency of GlobalSign TLS issuing CA rotation to every quarter in 2022. Our legacy platforms are expected to be fully deprecated for TLS in 2023.
We have generated the first set of CA that will follow this rotational scheme:
- GlobalSign Atlas R3 DV TLS CA 2020-12
- GlobalSign Atlas R3 OV TLS CA 2020-12
- GlobalSign Atlas ECCR5 DV TLS CA 2020-12
- GlobalSign Atlas ECCR5 OV TLS CA 2020-12
They are currently being set-up in the Atlas platform and we will swap issuance to these new CA at the end of this month. The next swap will be in July 2021 and as of January 2022 we will swap TLS issuance every quarter.
Assignee | ||
Comment 18•3 years ago
|
||
Hi Ben - I propose we close this bug and continue the tracking of the SOC2/SOC3 report on non-performance of OCSP EKU signing in https://bugzilla.mozilla.org/show_bug.cgi?id=1649937 as that is the one our customers and other relying parties are watching in the context of this incident. Thank you.
Comment 19•3 years ago
|
||
I will close this, and we can track any additional remediation action in Bug #1649937.
Updated•2 years ago
|
Updated•1 year ago
|
Description
•