Error 0x1000060 unknown returned by certverifycertificatechainpolicy

Tools Used in this Troubleshooter:

This material is provided for informational purposes only. Microsoft makes no warranties, express or implied.

Overview

This document will help you in troubleshooting SSL issues related to IIS only. Client Certificates troubleshooting will not be covered in this document. Server Certificates are meant for Server Authentication and we will be dealing only with Server Certificates in this document.

If the Client certificates section is set to «Require» and then you run into issues, then please don’t refer this document. This is meant for troubleshooting SSL Server certificates issue only.

It is important to know that every certificate comprises of a public key (used for encryption) and a private key (used for decryption). The private key is known only to the server.

The default port for https is 443.

I am under the assumption the reader is well-versed in SSL Handshake and the Server Authentication process during the SSL handshake.

Description of the Secure Sockets Layer (SSL) Handshake:

Description of the Server Authentication Process during the SSL Handshake:

Scenarios

The following error message is seen while browsing the website over https:

The first thing that has to be checked is whether the website is accessible over http. If it is not, there likely is a separate issue not covered here. You will need to have the website working on http first before continuing with this troubleshooter.

Now let’s assume the website is accessible over http and we get the above error when trying to browse over https. The problem is seen because the SSL handshake failed and hence the error message was seen. There could be many reasons. We will follow a step-by-step approach to solve this problem.

Scenario 1

Check if the server certificate has the private key corresponding to it. Refer the below picture:

If private key is missing, then you need to get a certificate containing the private key, which is essentially a .PFX file. There is a command that we could try to run in order to associate the private key with the certificate:

If the association is successful, then you would see the following window:

Note: 1a 1f 94 8b 21 a2 99 36 77 a8 8e b2 3f 42 8c 7e 47 e3 d1 33 is the thumbprint of the certificate. Open the certificate and click on the details tab. Scroll down to find the thumbprint section. Select the thumbprint section and click on the text below. Do a «Ctrl+A» and then «Ctrl+C» to select and copy it. Below is a snapshot for your reference:

Note: This command doesn’t succeed always. If this fails, then you need to get a certificate containing the private key from the CA. The file extension for a certificate containing private key is .pfx.

Scenario 2

We went pass the first hurdle and now we have a server certificate containing the private key installed on the website. However, we still get the same error as above. The website is still not accessible over https.

The SSLDiag tool comes in handy here.

Windows Server 2003:

Download X64 ( https://www.microsoft.com/download/en/details.aspx?displaylang=en&id=5329 )

Download X86 ( https://www.microsoft.com/download/en/details.aspx?displaylang=en&id=674 )

For IIS 7 and IIS 7.5, use vijaysk’s SSL Diagnostics tool. Below is the link:

Install the tool and run it on the server. If you have a certificate containing private key and still not able to access the website, then you may want to run this tool or check the system event logs for SChannel related warnings/errors.

While running the SSLDiag tool you may get the following error:

You have a private key that corresponds to this certificate but CryptAcquireCertificatePrivateKey failed

There will also be a SChannel warning in the system event logs as shown below:

This event/error indicates that there was a problem acquiring certificate’s private key. So let’s try the below steps one by one:

Firstly, verify the permissions on the machinekeys folder as per the KB Article: https://support.microsoft.com/kb/278381. All the private keys are stored within the machinekeys folder, so we need to ensure that we have necessary permissions.

If the permissions are in place and if the issue is still not fixed. Then it must be a problem with the certificate. It may have been corrupted (You may see an error code of 0x8009001a in the SChannel event log).

We will test if the website works with a test certificate. Take a back-up of the existing certificate and then replace it with a self-signed certificate. Try accessing the website via https. If it works then the certificate used earlier was corrupted and it has to be replaced with a new working certificate.

Sometimes the problem may not be with the certificate but with the issuer. You may see the following error in SSLDiag:

CertVerifyCertificateChainPolicy will fail with CERT_E_UNTRUSTEDROOT (0x800b0109), if the root CA certificate is not trusted root.

To fix this add the CA’s certificate to the «Trusted Root CA» store under My computer account on the server.

You may also get the following error:

CertVerifyCertificateChainPolicy returned error -2146762480(0x800b0110).

If the above error is received then we need to check the usage type of the certificate. Open the certificate, click on the «Details» tab and then click on «Edit Properties…» button. Under General tab make sure «Enable all purposes for this certificate» is selected and most importantly «Server Authentication» should be present in the list.

Scenario 3

The first 2 steps check the integrity of the certificate. Once we have confirmed that there are no issues with the certificate, a big problem is solved. But, what if the website is still not accessible over https. Check the HTTPS bindings of the website and determine what port and IP it is listening on. You could run the following command to ensure no other process is listening on the SSL port used by the website.

If there is another process listening on that port then check why that process is consuming that port. Try changing the IP-Port combination to check if the website is accessible or not.

Scenario 4

By now we are sure that we have a proper working certificate installed on the website and there is no other process using the SSL port for this website. However, I still get «Page cannot be displayed» error while accessing over https. When a client connects and initiates an SSL negotiation, HTTP.sys looks in its SSL configuration for the «IP:Port» pair to which the client connected. The HTTP.sys SSL configuration must include a certificate hash and the name of the certificate store before the SSL negotiation will succeed. The problem may be with the HTTP.SYS SSL Listener.

The Certificate hash registered with HTTP.SYS may be NULL or it may contain invalid GUID. Execute the following from a command prompt:

httpcfg is part of Windows Support tools and is present on the installation disk. You could download it from here as well: https://www.microsoft.com/download/en/details.aspx?id=7911

Below is a sample of a working and non-working scenario:

Working scenario:

Non-working scenario:

The Hash value seen above is the Thumbprint of your SSL certificate. Notice, that the Guid is all zero in a non-working scenario. You may see the Hash either having some value or blank. Even if we remove the certificate from the web site, and then run «httpcfg query ssl», the website will still list Guid as all 0’s. If you see the GUID as «<0000. 000>, then there is a problem.

We need to remove this entry by running the command:

Delete any entries in the IP Listen list.

To determine whether any IP addresses are listed, open a command prompt, and then run the following command:

If the IP Listen list is empty, the command returns the following string:

If the command returns a list of IP addresses, remove each IP address in the list by using the following command:

restart IIS after this via command «net stop http /y»

Scenario 5

After all this if you are still unable to browse the website on https, then capture a network trace either from the client or server. Filter the trace by «SSL or TLS» to look at SSL traffic.

Below is a network trace snapshot of a non-working scenario:

Well, this is definitely now how you look at a network trace. You need to expand the frame details and see what protocol and cipher was chosen by the server. Select «Server Hello» from the description to get those details.

In the non-working scenario, the client was configured to use TLS 1.1 and TLS 1.2 only. However, the web server was IIS 6, which can support until TLS 1.0 and hence the handshake failed.

Do check the registry keys to determine what protocols are enabled or disabled. Here’s the path:

The «Enabled» DWORD should be set to «1». If «0» then the protocol is disabled.

For example, SSL 2.0 is disabled by default.

Scenario 6

If everything has been verified and if you are still running into issues accessing the website over https, then it most likely is some update which is causing the SSL handshake to fail.

Microsoft has released an update to the implementation of SSL in Windows:

There is potential for this update to impact customers using Internet Explorer, or using an application that uses Internet Explorer to perform HTTPS requests.

There were actually two changes made to address information disclosure vulnerability in SSL 3.0 / TLS 1.0. The MS12-006 update implements a new behavior in schannel.dll, which sends an extra record while using a common SSL chained-block cipher, when clients request that behavior. The other change was in Wininet.dll, part of the December Cumulative Update for Internet Explorer (MS11-099), so that IE will request the new behavior.

If a problem exists, it may manifest as a failure to connect to a server, or an incomplete request. Internet Explorer 9 is able to display an «Internet Explorer cannot display the webpage» error. Prior versions of IE may simply display a blank page.

Fiddler does not use the extra record when it captures and forwards HTTPS requests to the server. Therefore, if Fiddler is used to capture HTTPS traffic, the requests will succeed.

Registry keys

As documented in https://support.microsoft.com/kb/2643584, there is a SendExtraRecord registry value, which can:

• Globally disable the new SSL behavior
• Globally enable it, or
• (Default) enable it for SChannel clients that opt in to the new behavior.

Источник

Tools Used in this Troubleshooter:

This material is provided for informational purposes only. Microsoft makes no warranties, express or implied.

Overview

This document will help you in troubleshooting SSL issues related to IIS only. Client Certificates troubleshooting will not be covered in this document. Server Certificates are meant for Server Authentication and we will be dealing only with Server Certificates in this document.

If the Client certificates section is set to «Require» and then you run into issues, then please don’t refer this document. This is meant for troubleshooting SSL Server certificates issue only.

It is important to know that every certificate comprises of a public key (used for encryption) and a private key (used for decryption). The private key is known only to the server.

The default port for https is 443.

I am under the assumption the reader is well-versed in SSL Handshake and the Server Authentication process during the SSL handshake.

Description of the Secure Sockets Layer (SSL) Handshake:

Description of the Server Authentication Process during the SSL Handshake:

Scenarios

The following error message is seen while browsing the website over https:

The first thing that has to be checked is whether the website is accessible over http. If it is not, there likely is a separate issue not covered here. You will need to have the website working on http first before continuing with this troubleshooter.

Now let’s assume the website is accessible over http and we get the above error when trying to browse over https. The problem is seen because the SSL handshake failed and hence the error message was seen. There could be many reasons. We will follow a step-by-step approach to solve this problem.

Scenario 1

Check if the server certificate has the private key corresponding to it. Refer the below picture:

If private key is missing, then you need to get a certificate containing the private key, which is essentially a .PFX file. There is a command that we could try to run in order to associate the private key with the certificate:

If the association is successful, then you would see the following window:

Note: 1a 1f 94 8b 21 a2 99 36 77 a8 8e b2 3f 42 8c 7e 47 e3 d1 33 is the thumbprint of the certificate. Open the certificate and click on the details tab. Scroll down to find the thumbprint section. Select the thumbprint section and click on the text below. Do a «Ctrl+A» and then «Ctrl+C» to select and copy it. Below is a snapshot for your reference:

Note: This command doesn’t succeed always. If this fails, then you need to get a certificate containing the private key from the CA. The file extension for a certificate containing private key is .pfx.

Scenario 2

We went pass the first hurdle and now we have a server certificate containing the private key installed on the website. However, we still get the same error as above. The website is still not accessible over https.

The SSLDiag tool comes in handy here.

Windows Server 2003:

Download X64 ( https://www.microsoft.com/download/en/details.aspx?displaylang=en&id=5329 )

Download X86 ( https://www.microsoft.com/download/en/details.aspx?displaylang=en&id=674 )

For IIS 7 and IIS 7.5, use vijaysk’s SSL Diagnostics tool. Below is the link:

Install the tool and run it on the server. If you have a certificate containing private key and still not able to access the website, then you may want to run this tool or check the system event logs for SChannel related warnings/errors.

While running the SSLDiag tool you may get the following error:

You have a private key that corresponds to this certificate but CryptAcquireCertificatePrivateKey failed

There will also be a SChannel warning in the system event logs as shown below:

This event/error indicates that there was a problem acquiring certificate’s private key. So let’s try the below steps one by one:

Firstly, verify the permissions on the machinekeys folder as per the KB Article: https://support.microsoft.com/kb/278381. All the private keys are stored within the machinekeys folder, so we need to ensure that we have necessary permissions.

If the permissions are in place and if the issue is still not fixed. Then it must be a problem with the certificate. It may have been corrupted (You may see an error code of 0x8009001a in the SChannel event log).

We will test if the website works with a test certificate. Take a back-up of the existing certificate and then replace it with a self-signed certificate. Try accessing the website via https. If it works then the certificate used earlier was corrupted and it has to be replaced with a new working certificate.

Sometimes the problem may not be with the certificate but with the issuer. You may see the following error in SSLDiag:

CertVerifyCertificateChainPolicy will fail with CERT_E_UNTRUSTEDROOT (0x800b0109), if the root CA certificate is not trusted root.

To fix this add the CA’s certificate to the «Trusted Root CA» store under My computer account on the server.

You may also get the following error:

CertVerifyCertificateChainPolicy returned error -2146762480(0x800b0110).

If the above error is received then we need to check the usage type of the certificate. Open the certificate, click on the «Details» tab and then click on «Edit Properties…» button. Under General tab make sure «Enable all purposes for this certificate» is selected and most importantly «Server Authentication» should be present in the list.

Scenario 3

The first 2 steps check the integrity of the certificate. Once we have confirmed that there are no issues with the certificate, a big problem is solved. But, what if the website is still not accessible over https. Check the HTTPS bindings of the website and determine what port and IP it is listening on. You could run the following command to ensure no other process is listening on the SSL port used by the website.

If there is another process listening on that port then check why that process is consuming that port. Try changing the IP-Port combination to check if the website is accessible or not.

Scenario 4

By now we are sure that we have a proper working certificate installed on the website and there is no other process using the SSL port for this website. However, I still get «Page cannot be displayed» error while accessing over https. When a client connects and initiates an SSL negotiation, HTTP.sys looks in its SSL configuration for the «IP:Port» pair to which the client connected. The HTTP.sys SSL configuration must include a certificate hash and the name of the certificate store before the SSL negotiation will succeed. The problem may be with the HTTP.SYS SSL Listener.

The Certificate hash registered with HTTP.SYS may be NULL or it may contain invalid GUID. Execute the following from a command prompt:

httpcfg is part of Windows Support tools and is present on the installation disk. You could download it from here as well: https://www.microsoft.com/download/en/details.aspx?id=7911

Below is a sample of a working and non-working scenario:

Working scenario:

Non-working scenario:

The Hash value seen above is the Thumbprint of your SSL certificate. Notice, that the Guid is all zero in a non-working scenario. You may see the Hash either having some value or blank. Even if we remove the certificate from the web site, and then run «httpcfg query ssl», the website will still list Guid as all 0’s. If you see the GUID as «<0000. 000>, then there is a problem.

We need to remove this entry by running the command:

Delete any entries in the IP Listen list.

To determine whether any IP addresses are listed, open a command prompt, and then run the following command:

If the IP Listen list is empty, the command returns the following string:

If the command returns a list of IP addresses, remove each IP address in the list by using the following command:

restart IIS after this via command «net stop http /y»

Scenario 5

After all this if you are still unable to browse the website on https, then capture a network trace either from the client or server. Filter the trace by «SSL or TLS» to look at SSL traffic.

Below is a network trace snapshot of a non-working scenario:

Well, this is definitely now how you look at a network trace. You need to expand the frame details and see what protocol and cipher was chosen by the server. Select «Server Hello» from the description to get those details.

In the non-working scenario, the client was configured to use TLS 1.1 and TLS 1.2 only. However, the web server was IIS 6, which can support until TLS 1.0 and hence the handshake failed.

Do check the registry keys to determine what protocols are enabled or disabled. Here’s the path:

The «Enabled» DWORD should be set to «1». If «0» then the protocol is disabled.

For example, SSL 2.0 is disabled by default.

Scenario 6

If everything has been verified and if you are still running into issues accessing the website over https, then it most likely is some update which is causing the SSL handshake to fail.

Microsoft has released an update to the implementation of SSL in Windows:

There is potential for this update to impact customers using Internet Explorer, or using an application that uses Internet Explorer to perform HTTPS requests.

There were actually two changes made to address information disclosure vulnerability in SSL 3.0 / TLS 1.0. The MS12-006 update implements a new behavior in schannel.dll, which sends an extra record while using a common SSL chained-block cipher, when clients request that behavior. The other change was in Wininet.dll, part of the December Cumulative Update for Internet Explorer (MS11-099), so that IE will request the new behavior.

If a problem exists, it may manifest as a failure to connect to a server, or an incomplete request. Internet Explorer 9 is able to display an «Internet Explorer cannot display the webpage» error. Prior versions of IE may simply display a blank page.

Fiddler does not use the extra record when it captures and forwards HTTPS requests to the server. Therefore, if Fiddler is used to capture HTTPS traffic, the requests will succeed.

Registry keys

As documented in https://support.microsoft.com/kb/2643584, there is a SendExtraRecord registry value, which can:

• Globally disable the new SSL behavior
• Globally enable it, or
• (Default) enable it for SChannel clients that opt in to the new behavior.

Источник

1С и Linux

Пишу для себя, чтобы не забыть как делал. 95 % рабочее. На комментарии отвечаю, когда увижу.

суббота, 15 декабря 2018 г.

Доступ zakupki.gov.ru Not Secure

]# /opt/cprocsp/bin/amd64/csptestf -tlsc -server zakupki.gov.ru -v -nosave -savecert /tmp/t.p7b
[root@test-x64-centos7

]# /opt/cprocsp/bin/amd64/certmgr -list -file /tmp/t.p7b
=============================================================================
1——-
Issuer : E=uc_fk@roskazna.ru, S=г. Москва, INN=007710568760, OGRN=1047797019830, STREET=»улица Ильинка, дом 7″, L=Москва, C=RU, O=Федеральное казначейство, CN=Федеральное казначейство
Subject : INN=007710568760, OGRN=1047797019830, STREET=»ул. Проспект Мира, 105″, E=777@roskazna.ru, C=RU, S=г. Москва, L=Москва, O=Федеральное казначейство, CN=zakupki.gov.ru
Serial : 0x6F064FA71C24CD7E2CE6FAAEA927D8C7EC69A35F
SHA1 Hash : c369b560ce239beddb2fc12b4884dee1cfc923aa
SubjKeyID : 10c6d12e7cd886d022bcdfea4cbe10e32acf82bc
Signature Algorithm : ГОСТ Р 34.11/34.10-2001
PublicKey Algorithm : ГОСТ Р 34.10-2001 (512 bits)
Public key : 04 40 1b ce 0b 15 6a 4a 57 e0 1b d8 39 ee 86 83
32 70 ea db fd f5 39 c1 e6 de 3c c8 be 10 81 03
5d da 0b 3b 09 7a d3 0d 46 33 58 11 3b 20 94 99
fe 04 fe 8e e6 bc 32 53 ff 2e 71 10 8e e2 12 a1
52 cf
Not valid before : 21/12/2017 06:06:33 UTC
Not valid after : 21/03/2019 06:06:33 UTC
PrivateKey Link : No
Subject Alt Names
UPN : ▒㌴ (2.5.4.12)
URL : 0
CDP : http://crl.roskazna.ru/crl/ucfk.crl
CDP : http://crl.fsfk.local/crl/ucfk.crl
Extended Key Usage : 1.3.6.1.5.5.7.3.1
.

При наличии расширения Subject Alt Names RFC требует сверять имя сервера с ним, а не с полем CN. Поэтому CN=zakupki.gov.ru отброшен, а в Subject Alt Names хрень.

ХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХ
Andrey, [05.10.18 11:35]
Но мы сделали параметр для отключения требований RFC6125, который может подвергнуть систему опасности для атак с подменой сертификатов:
[root@test-x64-centos7

]# /opt/cprocsp/sbin/amd64/cpconfig -ini ‘configparameters’ -add long Rfc6125_NotStrict_ServerName_Check 1
[root@test-x64-centos7

]# /opt/cprocsp/bin/amd64/csptestf -tlsc -server zakupki.gov.ru -nosave
HDEContext expired: OK if file is completely downloaded
Reply status: HTTP/1.1 200 OK
1 connections, 586 bytes in 0.186 seconds;
Total: SYS: 0,020 sec USR: 0,140 sec UTC: 0,230 sec
[ErrorCode: 0x00000000]

# /opt/cprocsp/bin/amd64/csptest -keyset -verifycontext
CSP (Type:80) v4.0.9017 KC2 Release Ver:4.0.9944 OS:Linux CPU:AMD64 FastCode:READY:AVX.
AcquireContext: OK. HCRYPTPROV: 25072019
GetProvParam(PP_NAME): Crypto-Pro GOST R 34.10-2012 KC2 CSP
Total: SYS: 0,000 sec USR: 0,000 sec UTC: 0,010 sec
[ErrorCode: 0x00000000]

L K, [15.12.18 14:12]
[В ответ на Andrey]
root@astra:

# /opt/cprocsp/sbin/amd64/cpconfig -ini ‘configparameters’ -add long Rfc6125_NotStrict_ServerName_Check 1
root@astra:

# /opt/cprocsp/bin/amd64/csptestf -tlsc -server zakupki.gov.ru -nosave
HSECPKG_ATTR_PACKAGE_INFO not supported.
Error 0x800b010f (CERT_E_CN_NO_MATCH) returned by CertVerifyCertificateChainPolicy!
An error occurred in running the program.
/dailybuildsbranches/CSP_4_0/CSPbuild/CSP/samples/csptest/WebClient.c:764:Error authenticating server credentials!
Error number 0x800b010f (2148204815).
CN-имя сертификата не совпадает с полученным значением.
Total: SYS: 0,070 sec USR: 0,290 sec UTC: 1,720 sec
[ErrorCode: 0x800b010f]

# /opt/cprocsp/bin/amd64/csptest -keyset -verifycontext
CSP (Type:80) v4.0.9017 KC2 Release Ver:4.0.9944 OS:Linux CPU:AMD64 FastCode:READY:AVX.
AcquireContext: OK. HCRYPTPROV: 25072019
GetProvParam(PP_NAME): Crypto-Pro GOST R 34.10-2012 KC2 CSP
Total: SYS: 0,000 sec USR: 0,000 sec UTC: 0,010 sec
[ErrorCode: 0x00000000]

Какой релиз КриптоПро нужно ставить?

ХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХ
Andrey, [12.12.18 18:12]
Вышел финальный релиз КриптоПро CSP 4.0 R4 (сборка 4.0.9963 Abel):
https://www.cryptopro.ru/forum2/default.aspx?g=posts&m=97735#post97735
На дистрибутив получено положительное заключение, скоро будет сертификат.
В тестовых целях её можно установить из нашего репозитория:
https://www.cryptopro.ru/forum2/default.aspx?g=posts&t=12507

ХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХ
Andrey, [15.12.18 23:07]
[В ответ на L K]
В любом случае лучше ставить последний, на который сейчас получаем сертификат — 4.0.9963 Abel. В 9944 в этой команде должна помочь замена long на bool и 1 на true.

Проблема решилась установкой КриптоПро версии 4.0.9963 и действиями как написал Андрей, за что ему огромное спасибо.

$ /opt/cprocsp/bin/amd64/csptest -keyset -verifycontext
CSP (Type:80) v4.0.9019 KC2 Release Ver:4.0.9963 OS:Linux CPU:AMD64 FastCode:READY:AVX.
AcquireContext: OK. HCRYPTPROV: 26834115
GetProvParam(PP_NAME): Crypto-Pro GOST R 34.10-2012 KC2 CSP
Total: SYS: 0,000 sec USR: 0,000 sec UTC: 0,020 sec
[ErrorCode: 0x00000000]

$ sudo /opt/cprocsp/sbin/amd64/cpconfig -ini ‘configparameters’ -add long Rfc6125_NotStrict_ServerName_Check 1

$ sudo /opt/cprocsp/bin/amd64/csptestf -tlsc -server zakupki.gov.ru -nosave
HSECPKG_ATTR_PACKAGE_INFO not supported.
DContext expired: OK if file is completely downloaded
Reply status: HTTP/1.1 200 OK
1 connections, 589 bytes in 1.221 seconds;
Total: SYS: 0,040 sec USR: 0,320 sec UTC: 1,300 sec
[ErrorCode: 0x00000000]

Для 4.0.9944
user@astra:

$ /opt/cprocsp/bin/amd64/csptest -keyset -verifycontext
CSP (Type:80) v4.0.9017 KC2 Release Ver:4.0.9944 OS:Linux CPU:AMD64 FastCode:READY:AVX.
AcquireContext: OK. HCRYPTPROV: 20066515
GetProvParam(PP_NAME): Crypto-Pro GOST R 34.10-2012 KC2 CSP
Total: SYS: 0,000 sec USR: 0,000 sec UTC: 0,030 sec
[ErrorCode: 0x00000000]
user@astra:

$ sudo /opt/cprocsp/sbin/amd64/cpconfig -ini ‘configparameters’ -add bool Rfc6125_NotStrict_ServerName_Check true
user@astra:

Источник

Tools Used in this Troubleshooter:

This material is provided for informational purposes only. Microsoft makes no warranties, express or implied.

Overview

This document will help you in troubleshooting SSL issues related to IIS only. Client Certificates troubleshooting will not be covered in this document. Server Certificates are meant for Server Authentication and we will be dealing only with Server Certificates in this document.

If the Client certificates section is set to «Require» and then you run into issues, then please don’t refer this document. This is meant for troubleshooting SSL Server certificates issue only.

It is important to know that every certificate comprises of a public key (used for encryption) and a private key (used for decryption). The private key is known only to the server.

The default port for https is 443.

I am under the assumption the reader is well-versed in SSL Handshake and the Server Authentication process during the SSL handshake.

Description of the Secure Sockets Layer (SSL) Handshake:

Description of the Server Authentication Process during the SSL Handshake:

Scenarios

The following error message is seen while browsing the website over https:

The first thing that has to be checked is whether the website is accessible over http. If it is not, there likely is a separate issue not covered here. You will need to have the website working on http first before continuing with this troubleshooter.

Now let’s assume the website is accessible over http and we get the above error when trying to browse over https. The problem is seen because the SSL handshake failed and hence the error message was seen. There could be many reasons. We will follow a step-by-step approach to solve this problem.

Scenario 1

Check if the server certificate has the private key corresponding to it. Refer the below picture:

If private key is missing, then you need to get a certificate containing the private key, which is essentially a .PFX file. There is a command that we could try to run in order to associate the private key with the certificate:

If the association is successful, then you would see the following window:

Note: 1a 1f 94 8b 21 a2 99 36 77 a8 8e b2 3f 42 8c 7e 47 e3 d1 33 is the thumbprint of the certificate. Open the certificate and click on the details tab. Scroll down to find the thumbprint section. Select the thumbprint section and click on the text below. Do a «Ctrl+A» and then «Ctrl+C» to select and copy it. Below is a snapshot for your reference:

Note: This command doesn’t succeed always. If this fails, then you need to get a certificate containing the private key from the CA. The file extension for a certificate containing private key is .pfx.

Scenario 2

We went pass the first hurdle and now we have a server certificate containing the private key installed on the website. However, we still get the same error as above. The website is still not accessible over https.

The SSLDiag tool comes in handy here.

Windows Server 2003:

Download X64 ( https://www.microsoft.com/download/en/details.aspx?displaylang=en&id=5329 )

Download X86 ( https://www.microsoft.com/download/en/details.aspx?displaylang=en&id=674 )

For IIS 7 and IIS 7.5, use vijaysk’s SSL Diagnostics tool. Below is the link:

Install the tool and run it on the server. If you have a certificate containing private key and still not able to access the website, then you may want to run this tool or check the system event logs for SChannel related warnings/errors.

While running the SSLDiag tool you may get the following error:

You have a private key that corresponds to this certificate but CryptAcquireCertificatePrivateKey failed

There will also be a SChannel warning in the system event logs as shown below:

This event/error indicates that there was a problem acquiring certificate’s private key. So let’s try the below steps one by one:

Firstly, verify the permissions on the machinekeys folder as per the KB Article: https://support.microsoft.com/kb/278381. All the private keys are stored within the machinekeys folder, so we need to ensure that we have necessary permissions.

If the permissions are in place and if the issue is still not fixed. Then it must be a problem with the certificate. It may have been corrupted (You may see an error code of 0x8009001a in the SChannel event log).

We will test if the website works with a test certificate. Take a back-up of the existing certificate and then replace it with a self-signed certificate. Try accessing the website via https. If it works then the certificate used earlier was corrupted and it has to be replaced with a new working certificate.

Sometimes the problem may not be with the certificate but with the issuer. You may see the following error in SSLDiag:

CertVerifyCertificateChainPolicy will fail with CERT_E_UNTRUSTEDROOT (0x800b0109), if the root CA certificate is not trusted root.

To fix this add the CA’s certificate to the «Trusted Root CA» store under My computer account on the server.

You may also get the following error:

CertVerifyCertificateChainPolicy returned error -2146762480(0x800b0110).

If the above error is received then we need to check the usage type of the certificate. Open the certificate, click on the «Details» tab and then click on «Edit Properties…» button. Under General tab make sure «Enable all purposes for this certificate» is selected and most importantly «Server Authentication» should be present in the list.

Scenario 3

The first 2 steps check the integrity of the certificate. Once we have confirmed that there are no issues with the certificate, a big problem is solved. But, what if the website is still not accessible over https. Check the HTTPS bindings of the website and determine what port and IP it is listening on. You could run the following command to ensure no other process is listening on the SSL port used by the website.

If there is another process listening on that port then check why that process is consuming that port. Try changing the IP-Port combination to check if the website is accessible or not.

Scenario 4

By now we are sure that we have a proper working certificate installed on the website and there is no other process using the SSL port for this website. However, I still get «Page cannot be displayed» error while accessing over https. When a client connects and initiates an SSL negotiation, HTTP.sys looks in its SSL configuration for the «IP:Port» pair to which the client connected. The HTTP.sys SSL configuration must include a certificate hash and the name of the certificate store before the SSL negotiation will succeed. The problem may be with the HTTP.SYS SSL Listener.

The Certificate hash registered with HTTP.SYS may be NULL or it may contain invalid GUID. Execute the following from a command prompt:

httpcfg is part of Windows Support tools and is present on the installation disk. You could download it from here as well: https://www.microsoft.com/download/en/details.aspx?id=7911

Below is a sample of a working and non-working scenario:

Working scenario:

Non-working scenario:

The Hash value seen above is the Thumbprint of your SSL certificate. Notice, that the Guid is all zero in a non-working scenario. You may see the Hash either having some value or blank. Even if we remove the certificate from the web site, and then run «httpcfg query ssl», the website will still list Guid as all 0’s. If you see the GUID as «<0000. 000>, then there is a problem.

We need to remove this entry by running the command:

Delete any entries in the IP Listen list.

To determine whether any IP addresses are listed, open a command prompt, and then run the following command:

If the IP Listen list is empty, the command returns the following string:

If the command returns a list of IP addresses, remove each IP address in the list by using the following command:

restart IIS after this via command «net stop http /y»

Scenario 5

After all this if you are still unable to browse the website on https, then capture a network trace either from the client or server. Filter the trace by «SSL or TLS» to look at SSL traffic.

Below is a network trace snapshot of a non-working scenario:

Well, this is definitely now how you look at a network trace. You need to expand the frame details and see what protocol and cipher was chosen by the server. Select «Server Hello» from the description to get those details.

In the non-working scenario, the client was configured to use TLS 1.1 and TLS 1.2 only. However, the web server was IIS 6, which can support until TLS 1.0 and hence the handshake failed.

Do check the registry keys to determine what protocols are enabled or disabled. Here’s the path:

The «Enabled» DWORD should be set to «1». If «0» then the protocol is disabled.

For example, SSL 2.0 is disabled by default.

Scenario 6

If everything has been verified and if you are still running into issues accessing the website over https, then it most likely is some update which is causing the SSL handshake to fail.

Microsoft has released an update to the implementation of SSL in Windows:

There is potential for this update to impact customers using Internet Explorer, or using an application that uses Internet Explorer to perform HTTPS requests.

There were actually two changes made to address information disclosure vulnerability in SSL 3.0 / TLS 1.0. The MS12-006 update implements a new behavior in schannel.dll, which sends an extra record while using a common SSL chained-block cipher, when clients request that behavior. The other change was in Wininet.dll, part of the December Cumulative Update for Internet Explorer (MS11-099), so that IE will request the new behavior.

If a problem exists, it may manifest as a failure to connect to a server, or an incomplete request. Internet Explorer 9 is able to display an «Internet Explorer cannot display the webpage» error. Prior versions of IE may simply display a blank page.

Fiddler does not use the extra record when it captures and forwards HTTPS requests to the server. Therefore, if Fiddler is used to capture HTTPS traffic, the requests will succeed.

Registry keys

As documented in https://support.microsoft.com/kb/2643584, there is a SendExtraRecord registry value, which can:

• Globally disable the new SSL behavior
• Globally enable it, or
• (Default) enable it for SChannel clients that opt in to the new behavior.

Источник

Boris7772	#1 Оставлено : 24 ноября 2020 г. 12:47:18(UTC)
Статус: Новичок Группы: Участники Зарегистрирован: 21.11.2020(UTC) Сообщений: 3	Добрый день, Пытаюсь настроить взаимодейтсвие с ФинЦЕРТ через stunnel установил согласно регламенту ФинЦЕРТ сертификаты из архива: http://cbr.ru/StaticHtml/File/14408/ASOI_docs.zip Приложил лог с ошибкой и конфигурацию stunnel. Использую stunnel из CSP 4.0.9963 Конфигурация stunnel setgid = stunnel setuid = stunnel pid=/var/opt/cprocsp/tmp/stunnel_cli.pid output=/var/opt/cprocsp/tmp/stunnel_cli.log socket = l:TCP_NODELAY=1 socket = r:TCP_NODELAY=1 debug = 7 [portal] client=yes accept = localhost:8181 connect = portal.fincert.cbr.ru:443 verify=2 Лог с проблемой: Цитата: 2020.11.24 12:30:44 LOG3[1123:140439079679744]: Error 0x8009030e returned by VerifyCertChain 2020.11.24 12:30:44 LOG3[1123:140439079679744]: **** Error 0x8009030e authenticating server credentials! 2020.11.24 12:30:44 LOG5[1123:140439079679744]: Connection reset: 0 bytes sent to SSL, 0 bytes sent to socket 2020.11.24 12:30:44 LOG7[1123:140439079679744]: free Buffers 2020.11.24 12:30:44 LOG7[1123:140439079679744]: delete c->hContext 2020.11.24 12:30:44 LOG7[1123:140439079679744]: delete c->hClientCreds 2020.11.24 12:30:44 LOG5[1123:140439079679744]: incomp_mess = 0, extra_data = 0 2020.11.24 12:30:44 LOG7[1123:140439079679744]: portal finished (0 left) 2020.11.24 12:30:54 LOG7[1123:140439079688000]: portal accepted FD=8 from ::1:48772 2020.11.24 12:30:54 LOG7[1123:140439079679744]: client start 2020.11.24 12:30:54 LOG7[1123:140439079679744]: portal started 2020.11.24 12:30:54 LOG7[1123:140439079679744]: FD 8 in non-blocking mode 2020.11.24 12:30:54 LOG7[1123:140439079679744]: TCP_NODELAY option set on local socket 2020.11.24 12:30:54 LOG7[1123:140439079679744]: FD 10 in non-blocking mode 2020.11.24 12:30:54 LOG7[1123:140439079679744]: FD 11 in non-blocking mode 2020.11.24 12:30:54 LOG7[1123:140439079679744]: Connection from ::1:48772 permitted by libwrap 2020.11.24 12:30:54 LOG5[1123:140439079679744]: portal connected from ::1:48772 2020.11.24 12:30:54 LOG7[1123:140439079688000]: Cleaning up the signal pipe 2020.11.24 12:30:54 LOG6[1123:140439079688000]: Child process 1739 finished with code 0 2020.11.24 12:30:54 LOG7[1123:140439079679744]: FD 16 in non-blocking mode 2020.11.24 12:30:54 LOG7[1123:140439079679744]: portal connecting 2020.11.24 12:30:54 LOG7[1123:140439079679744]: connect_wait: waiting 10 seconds 2020.11.24 12:30:54 LOG7[1123:140439079679744]: connect_wait: connected 2020.11.24 12:30:54 LOG7[1123:140439079679744]: Remote FD=16 initialized 2020.11.24 12:30:54 LOG7[1123:140439079679744]: TCP_NODELAY option set on remote socket 2020.11.24 12:30:54 LOG7[1123:140439079679744]: start SSPI connect 2020.11.24 12:30:54 LOG3[1123:140439079679744]: Credentials complete 2020.11.24 12:30:54 LOG7[1123:140439079679744]: 113 bytes of handshake data sent 2020.11.24 12:30:54 LOG5[1123:140439079679744]: 2018 bytes of handshake(in handshake loop) data received. 2020.11.24 12:30:54 LOG5[1123:140439079679744]: 215 bytes of handshake data sent 2020.11.24 12:30:54 LOG5[1123:140439079679744]: 31 bytes of handshake(in handshake loop) data received. 2020.11.24 12:30:54 LOG5[1123:140439079679744]: Handshake was successful 2020.11.24 12:30:54 LOG5[1123:140439079679744]: PerformClientHandshake finish 2020.11.24 12:30:54 LOG5[1123:140439079679744]: Server subject: 1.2.643.100.111=»#0C818AD0A1D0A0D095D094D0A1D0A2D092D09E20D09AD0A0D098D09FD0A2D09ED093D0A0D090D0A4D098D0A7D095D0A1D09AD09ED09920D097D090D0A9D098D0A2D0AB20D098D09DD0A4D09ED0A0D09CD090D0A6D098D0982022D09AD09ED09DD0A2D098D09DD095D09DD0A220544C532DD0A1D095D0A0D092D095D0A02220D092D095D0A0D0A1D098D0AF2032″, C=RU, L=Москва, O=CBR, OU=Центр мониторинга и реагирования на компьютерные атаки ЦБ РФ, STREET=»ул.Неглинная, д.12″, CN=portal.fincert.cbr.ru, E=info_fincert@cbr.ru 2020.11.24 12:30:54 LOG5[1123:140439079679744]: Server issuer: DC=ru, DC=cbr, DC=region, OU=PKI, OU=GUBZI, CN=ROOTsvc-CA 2020.11.24 12:30:54 LOG3[1123:140439079679744]: Error 0x10000 ((unknown)) returned by CertVerifyCertificateChainPolicy! 2020.11.24 12:30:54 LOG3[1123:140439079679744]: Error 0x10000 when validate certificate 2020.11.24 12:30:54 LOG3[1123:140439079679744]: Error 0x8009030e returned by VerifyCertChain 2020.11.24 12:30:54 LOG3[1123:140439079679744]: **** Error 0x8009030e authenticating server credentials! 2020.11.24 12:30:54 LOG5[1123:140439079679744]: Connection reset: 0 bytes sent to SSL, 0 bytes sent to socket 2020.11.24 12:30:54 LOG7[1123:140439079679744]: free Buffers 2020.11.24 12:30:54 LOG7[1123:140439079679744]: delete c->hContext 2020.11.24 12:30:54 LOG7[1123:140439079679744]: delete c->hClientCreds 2020.11.24 12:30:54 LOG5[1123:140439079679744]: incomp_mess = 0, extra_data = 0 2020.11.24 12:30:54 LOG7[1123:140439079679744]: portal finished (0 left) Подскажите, пожалуйста, с чем связана данная проблема и как ее можно решить? Отредактировано пользователем 24 ноября 2020 г. 18:01:23(UTC)  | Причина: Не указана

Михаил Селезнёв	#2 Оставлено : 24 ноября 2020 г. 17:52:02(UTC)
Статус: Сотрудник Группы: Участники Зарегистрирован: 05.04.2017(UTC) Сообщений: 354 Сказал «Спасибо»: 3 раз Поблагодарили: 49 раз в 49 постах	Добрый день! Приложенные файлы пустые. Прошу запаковать их в архив.

Boris7772	#3 Оставлено : 24 ноября 2020 г. 18:02:28(UTC)
Статус: Новичок Группы: Участники Зарегистрирован: 21.11.2020(UTC) Сообщений: 3	Автор: Михаил Селезнёв Добрый день! Приложенные файлы пустые. Прошу запаковать их в архив. Что-то не удалось загрузить файлы нормально, опубликовал лог и конфигурацию в посте

Nikolay Batischev	#4 Оставлено : 24 ноября 2020 г. 18:13:58(UTC)
Статус: Сотрудник Группы: Участники Зарегистрирован: 06.11.2013(UTC) Сообщений: 70 Откуда: Тирана Сказал «Спасибо»: 3 раз Поблагодарили: 11 раз в 11 постах	Добрый день! Нужно установить корневой сертификат от сертификата в доверенное хранилище. Или в конфиге использовать verify =0

Boris7772	#5 Оставлено : 24 ноября 2020 г. 18:26:15(UTC)
Статус: Новичок Группы: Участники Зарегистрирован: 21.11.2020(UTC) Сообщений: 3	Автор: Nikolay Batischev Добрый день! Нужно установить корневой сертификат от сертификата в доверенное хранилище. Или в конфиге использовать verify =0 Установил этот параметр в 0, ошибка ушла спасибо, запросы к сервису стали отбиваться с 421 статусом, это я так понимаю уже внешняя проблема? Лог при этом следующий: 2020.11.24 18:28:35 LOG7[23480:139871409022784]: portal accepted FD=8 from 192.168.24.98:59128 2020.11.24 18:28:35 LOG7[23480:139871409014528]: client start 2020.11.24 18:28:35 LOG7[23480:139871409014528]: portal started 2020.11.24 18:28:35 LOG7[23480:139871409014528]: FD 8 in non-blocking mode 2020.11.24 18:28:35 LOG7[23480:139871409014528]: TCP_NODELAY option set on local socket 2020.11.24 18:28:35 LOG7[23480:139871409014528]: FD 9 in non-blocking mode 2020.11.24 18:28:35 LOG7[23480:139871409014528]: FD 11 in non-blocking mode 2020.11.24 18:28:35 LOG7[23480:139871409014528]: Connection from 192.168.24.98:59128 permitted by libwrap 2020.11.24 18:28:35 LOG5[23480:139871409014528]: portal connected from 192.168.24.98:59128 2020.11.24 18:28:35 LOG7[23480:139871409022784]: Cleaning up the signal pipe 2020.11.24 18:28:35 LOG6[23480:139871409022784]: Child process 24405 finished with code 0 2020.11.24 18:28:35 LOG7[23480:139871409014528]: FD 15 in non-blocking mode 2020.11.24 18:28:35 LOG7[23480:139871409014528]: portal connecting 2020.11.24 18:28:35 LOG7[23480:139871409014528]: connect_wait: waiting 10 seconds 2020.11.24 18:28:35 LOG7[23480:139871409014528]: connect_wait: connected 2020.11.24 18:28:35 LOG7[23480:139871409014528]: Remote FD=15 initialized 2020.11.24 18:28:35 LOG7[23480:139871409014528]: TCP_NODELAY option set on remote socket 2020.11.24 18:28:35 LOG7[23480:139871409014528]: start SSPI connect 2020.11.24 18:28:35 LOG3[23480:139871409014528]: Credentials complete 2020.11.24 18:28:35 LOG7[23480:139871409014528]: 118 bytes of handshake data sent 2020.11.24 18:28:35 LOG5[23480:139871409014528]: 1975 bytes of handshake(in handshake loop) data received. 2020.11.24 18:28:35 LOG5[23480:139871409014528]: 215 bytes of handshake data sent 2020.11.24 18:28:35 LOG5[23480:139871409014528]: 31 bytes of handshake(in handshake loop) data received. 2020.11.24 18:28:35 LOG5[23480:139871409014528]: Handshake was successful 2020.11.24 18:28:35 LOG5[23480:139871409014528]: PerformClientHandshake finish 2020.11.24 18:28:35 LOG5[23480:139871409014528]: Verify_level = 0, skipping Server certificate verification 2020.11.24 18:28:35 LOG7[23480:139871409014528]: add ssl read socket to pool 2020.11.24 18:28:35 LOG7[23480:139871409014528]: ssl_rd = 1, c->ssl_ptr = 0,c->sock_ptr=0,want_rd = 0 2020.11.24 18:28:35 LOG7[23480:139871409014528]: Enter pool section on transfer 2020.11.24 18:28:35 LOG7[23480:139871409014528]: data reciev from socket = 116 2020.11.24 18:28:35 LOG7[23480:139871409014528]: add ssl read socket to pool 2020.11.24 18:28:35 LOG7[23480:139871409014528]: ssl_rd = 1, c->ssl_ptr = 0,c->sock_ptr=116,want_rd = 0 2020.11.24 18:28:35 LOG7[23480:139871409014528]: Enter pool section on transfer 2020.11.24 18:28:35 LOG5[23480:139871409014528]: SSPI_write start 2020.11.24 18:28:35 LOG7[23480:139871409014528]: SSPI_write data is GET 2020.11.24 18:28:35 LOG7[23480:139871409014528]: send all data after encrypt 2020.11.24 18:28:35 LOG7[23480:139871409014528]: data send to ssl_socket =116 2020.11.24 18:28:35 LOG7[23480:139871409014528]: add ssl read socket to pool 2020.11.24 18:28:35 LOG7[23480:139871409014528]: ssl_rd = 1, c->ssl_ptr = 0,c->sock_ptr=0,want_rd = 0 2020.11.24 18:28:35 LOG7[23480:139871409014528]: Enter pool section on transfer 2020.11.24 18:28:35 LOG7[23480:139871409014528]: SSPI_read start 2020.11.24 18:28:35 LOG7[23480:139871409014528]: recv ok on SSPI_read err= 324 2020.11.24 18:28:35 LOG5[23480:139871409014528]: Received 324 bytes from ssl socket 2020.11.24 18:28:35 LOG7[23480:139871409014528]: SSPI_read data in ssl_buff is HTTP 2020.11.24 18:28:35 LOG7[23480:139871409014528]: data read from ssl_sock =315 2020.11.24 18:28:35 LOG7[23480:139871409014528]: add ssl read socket to pool 2020.11.24 18:28:35 LOG7[23480:139871409014528]: ssl_rd = 1, c->ssl_ptr = 315,c->sock_ptr=0,want_rd = 0 2020.11.24 18:28:35 LOG7[23480:139871409014528]: add write socket to poll 2020.11.24 18:28:35 LOG7[23480:139871409014528]: Enter pool section on transfer 2020.11.24 18:28:35 LOG7[23480:139871409014528]: data send to socket = 315 2020.11.24 18:28:35 LOG7[23480:139871409014528]: SSPI_read start 2020.11.24 18:28:35 LOG7[23480:139871409014528]: recv ok on SSPI_read err= 0 2020.11.24 18:28:35 LOG3[23480:139871409014528]: recv return 0 and ask more but there is not complete data for decrypt. Peer disconnected? 2020.11.24 18:28:35 LOG5[23480:139871409014528]: SSPI_read: read socket closed 2020.11.24 18:28:35 LOG7[23480:139871409014528]: Socket write shutdown 2020.11.24 18:28:35 LOG7[23480:139871409014528]: c->ssl_ptr = 0 2020.11.24 18:28:35 LOG7[23480:139871409014528]: Enter pool section on transfer 2020.11.24 18:28:35 LOG3[23480:139871409014528]: readsocket: Connection reset by peer (104) 2020.11.24 18:28:35 LOG5[23480:139871409014528]: Connection reset: 116 bytes sent to SSL, 315 bytes sent to socket 2020.11.24 18:28:35 LOG7[23480:139871409014528]: free Buffers 2020.11.24 18:28:35 LOG7[23480:139871409014528]: delete c->hContext 2020.11.24 18:28:35 LOG7[23480:139871409014528]: delete c->hClientCreds 2020.11.24 18:28:35 LOG5[23480:139871409014528]: incomp_mess = 2, extra_data = 0 2020.11.24 18:28:35 LOG7[23480:139871409014528]: portal finished (0 left) Отредактировано пользователем 24 ноября 2020 г. 18:29:32(UTC)  | Причина: Не указана

Быстрый переход
 

Вы не можете создавать новые темы в этом форуме.

Вы не можете отвечать в этом форуме.

Вы не можете удалять Ваши сообщения в этом форуме.

Вы не можете редактировать Ваши сообщения в этом форуме.

Вы не можете создавать опросы в этом форуме.

Вы не можете голосовать в этом форуме.

Иван Бахтин	#1 Оставлено : 4 апреля 2022 г. 15:11:58(UTC)
Статус: Участник Группы: Участники Зарегистрирован: 24.02.2021(UTC) Сообщений: 18 Откуда: Сыктывкар Сказал(а) «Спасибо»: 5 раз	Здравствуйте! Пытаюсь наладить интеграцию с ГИС ЖКХ через их API + stunnel. И выходит ошибка Error 0x8009030e returned by VerifyCertChain. Установленный сертификат: Настройки stunnel.conf: Вывод log stunnel: Так же хочу сказать, что проверил сертификаты на совпадение (т.е. установленный у меня и передаваемый с сервера ГИС ЖКХ) утилитой /opt/cprocsp/bin/amd64/csptest, сертификаты совпадают. Подскажите, пожалуйста, где я допустил ошибку и как её исправить? Спасибо!

two_oceans	#2 Оставлено : 5 апреля 2022 г. 7:20:56(UTC)
Статус: Эксперт Группы: Участники Зарегистрирован: 05.03.2015(UTC) Сообщений: 1,598 Откуда: Иркутская область Сказал(а) «Спасибо»: 110 раз Поблагодарили: 388 раз в 363 постах	Добрый день. Вижу 2 варианта исправления: 1) verify=0 и не задаваться таким неблагодарным делом как проверка сертификата, выданного тестовым УЦ. Имитировать такое сложное API как у ГИС ЖКХ — сомневаюсь, что хакеры будут. Если работает, значит это подлинная ГИС ЖКХ. Серьезно, добавлять тестовый УЦ в доверенные представляет намного больший риск чем игнорировать проверку сертификата сервера. Отключение проверки действует только на конкретное соединение, а тестовый УЦ неизвестно где потом вылезет. Если ГИС ЖКХ реально хотят безопасности, то они должны серверный сертификат получить у Минцифры, а не вот эту подделку для галочки. Для тестовых соединений так точно стоит ставить verify=0, так как там реальных персональных данных не передается. 2) в строке CAFile по идее должен быть не сам сертификат сервера, а список сертификатов УЦ, выдавшего сертификат серверу. В случае тестового УЦ КриптоПро это обычно 2 сертификата от корневого и промежуточного УЦ. Обычно это выглядит как: Код: ===== BEGIN TRUSTED CERTIFICATE ===== сертификат 1 ===== END TRUSTED CERTIFICATE ===== ===== BEGIN TRUSTED CERTIFICATE ===== сертификат 2 ===== END TRUSTED CERTIFICATE ===== заголовки могут немного отличаться смотря какая версия stunnel и openssl Отредактировано пользователем 5 апреля 2022 г. 7:27:39(UTC)  | Причина: Не указана
1 пользователь поблагодарил two_oceans за этот пост.	Иван Бахтин оставлено 05.04.2022(UTC)

Иван Бахтин	#3 Оставлено : 5 апреля 2022 г. 11:51:18(UTC)
Статус: Участник Группы: Участники Зарегистрирован: 24.02.2021(UTC) Сообщений: 18 Откуда: Сыктывкар Сказал(а) «Спасибо»: 5 раз	Автор: two_oceans Добрый день. Вижу 2 варианта исправления: 1) verify=0 и не задаваться таким неблагодарным делом как проверка сертификата, выданного тестовым УЦ. Имитировать такое сложное API как у ГИС ЖКХ — сомневаюсь, что хакеры будут. Если работает, значит это подлинная ГИС ЖКХ. Серьезно, добавлять тестовый УЦ в доверенные представляет намного больший риск чем игнорировать проверку сертификата сервера. Отключение проверки действует только на конкретное соединение, а тестовый УЦ неизвестно где потом вылезет. Если ГИС ЖКХ реально хотят безопасности, то они должны серверный сертификат получить у Минцифры, а не вот эту подделку для галочки. Для тестовых соединений так точно стоит ставить verify=0, так как там реальных персональных данных не передается. 2) в строке CAFile по идее должен быть не сам сертификат сервера, а список сертификатов УЦ, выдавшего сертификат серверу. В случае тестового УЦ КриптоПро это обычно 2 сертификата от корневого и промежуточного УЦ. Обычно это выглядит как: Код: ===== BEGIN TRUSTED CERTIFICATE ===== сертификат 1 ===== END TRUSTED CERTIFICATE ===== ===== BEGIN TRUSTED CERTIFICATE ===== сертификат 2 ===== END TRUSTED CERTIFICATE ===== заголовки могут немного отличаться смотря какая версия stunnel и openssl Большое спасибо! С verify=0 запросы проходят. Очень надеюсь, что на боевом сервере с verify=2 всё будет работать.

Быстрый переход
 

Вы не можете создавать новые темы в этом форуме.

Вы не можете отвечать в этом форуме.

Вы не можете удалять Ваши сообщения в этом форуме.

Вы не можете редактировать Ваши сообщения в этом форуме.

Вы не можете создавать опросы в этом форуме.

Вы не можете голосовать в этом форуме.

Всем доброго дня.
Недавно ко мне обратился знакомый с вопросом по реализации шифрованного по алгоритму ГОСТ TLS-канала для работы с веб-сервисами некоторых БКИ.
Немного помыслив, вспомнил, что есть такая замечательная штука как stunnel, которая может TLS как в режиме сервера, так и в режиме клиента. Плюс к этому компания КриптоПро сделала форк stunnel с оптимизацией под российские криптоалгоритмы. В итоге родился небольшой гайд по реализации простого и бюджетного варианта, которым и хочу поделиться с вами.

В качестве примера приводится настройка stunnel в качестве TLS клиента:
1) к серверу НБКИ без авторизации пользователя;
2) к тестовому УЦ КриптоПро с авторизацией пользователя по сертификату.
В данном материале использовались OC Debian 8.3 x86 и КриптоПро CSP 4.0.9708 (Nechaev) от 01.03.2016 (stunnel включен).

1. Установка Крипто-Про CSP и stunnel

Скачиваем архив CSP с сайта КриптоПро и распаковываем

tar zxvf linux-ia32_deb.tgz
cd linux-ia32_deb

Установливаем минимально необходимый набор пакетов

sudo apt-get install lsb-core
sudo dpkg -i lsb-cprocsp-base_4.0.0-4_all.deb
sudo dpkg -i lsb-cprocsp-rdr_4.0.0-4_i386.deb
sudo dpkg -i lsb-cprocsp-capilite_4.0.0-4_i386.deb
sudo dpkg -i lsb-cprocsp-kc1_4.0.0-4_i386.deb
sudo dpkg -i cprocsp-stunnel_4.0.0-4_i386.deb

2. Импорт корневых сертификатов

Импортируем корневые сертификаты КриптоПро в локальное хранилище сертификатов.
Нам понадобятся два сертификата:
— корневой сертификат УЦ Крипто-Про. Данным сертификатом подписан сертификат веб-сервера НБКИ;
— корневой сертификат тестового УЦ КриптоПро. Данным сертификатом подписан сертификат веб-сервера тестового УЦ КриптоПро и тестовый сертификат нашего пользователя.

/opt/cprocsp/bin/ia32/certmgr -inst -file <home>/cert/ca_cryptopro_4aad6f_der.cer -store ROOT
/opt/cprocsp/bin/ia32/certmgr -inst -file <home>/cert/ca_test_cryptopro_4487da_der.cer -store ROOT

После импортирования полезно просмотреть список установленных корневых сертификатов

/opt/cprocsp/bin/ia32/certmgr -list -store ROOT

3. Копирование ключевого контейнера пользователя

Наверное, это самая запутанная и самая важная часть данного руководства.
Сначала немного теории «на пальцах».
В протоколе TLS предусмотрена односторонняя и двусторонняя аутентификация. Под односторонней чаще всего понимается аутентификация веб-сервера, когда сервер доказывает клиенту что действительно владеет закрытым ключом предъявленного цифрового сертификата. Под двусторонней аутентификацией сервер и клиент доказывают это друг другу.
В КриптоПро ключевая информация (ключи шифрования) хранятся в контейнере (с точки зрения файловой системы — это каталог с несколькими файлами). В качестве носителя может выступать (чаще всего) дискета, флешка, токен, ветка реестра (Windows) или каталог (Linux). И если клиент или сервер проходят взаимную аутентификацию (доказывают владение ключами), то они должны иметь не только цифровой сертификат с открытым ключом, но и сам контейнер с ключами.
Опустим процесс создания тестового ключа в тестовом УЦ КриптоПро и начем с того момента, что у нас есть ключевой контейнер с именем «test» на флешке.
Монтируем флешку

sudo mount /dev/sdb1 /media/sdb1/

Копируем контейнер с флешки на HDD
Есть два способа: ручной и с помощью утилиты. Я скопировал вручную

cp -r /media/sdb1/test.000/ /var/opt/cprocsp/keys/stunnel/

Проверим сертификат в контейнере

certmgr -list -cont '\.HDIMAGEtest'

Проверим работоспособность ключей

csptest -keyset -check -cont '\.HDIMAGEtest'

Экспортируем сертификат пользователя в файл (он еще понадобится)

certmgr -export -dest test_user_7a8110.cer -cont '\.HDIMAGEtest'

Делаем привязку данного сертификата пользователя к нашему контейнеру

certmgr -inst -file test_user_7a8110.cer -cont '\.HDIMAGEtest'

Проверяем локальное хранилище сертификатов пользователя

certmgr -list -store My

Проверяем локальное хранилище сертификатов пользователя

PrivateKey Link     : Yes
Container           : HDIMAGE\test.0002EF8

Создаем файл конфигурации с нужными нам настройками для stunnel 

sudo mkdir /etc/stunnel
sudo vim /etc/stunnel/stunnel.conf

Содержимое файла конфигурации:

setgid = stunnel
setuid = stunnel
pid=/var/opt/cprocsp/tmp/stunnel_cli.pid
output=/var/opt/cprocsp/tmp/stunnel_cli.log
socket = l:TCP_NODELAY=1
socket = r:TCP_NODELAY=1
debug = 3
client = yes

[stunnel.nbki]
accept = <IP address>:8181
connect = icrs.nbki.ru:443
verify=2

[stunnel.test_uc]
accept = <IP address>:8282 
connect = www.cryptopro.ru:5555
client = /etc/stunnel/test_user_7a8110.cer
verify=2

Запускаем процесс

/opt/cprocsp/sbin/amd64/stunnel_thread /etc/stunnel/stunnel.conf

Проверяем результат

5. Автоматический запуск Stunnel

Когда все заработало должным образом, то было бы неплохо автоматизировать запуск stunnel при загрузке ОС.
Создаем скрипт запуска. Также будут доступны команды управления start, stop, restart.
Пример команды для перезагрузки stunnel:

sudo service stunnel_thread restart

Создаем скрипт

sudo vim /etc/init.d/stunnel_thread

и помещаем в него следующий текст:

#!/bin/sh
### BEGIN INIT INFO
# Provides:          stunnel
# Default-Start:     3 5
# Default-Stop:      0 1 6
# Required-Start:    $remote_fs $syslog
# Required-Stop:     $remote_fs $syslog
# Short-Description: Start daemon at boot time
# Description:       Enable stunnel by daemon.
### END INIT INFO

PATH_DAEMON=/opt/cprocsp/sbin/ia32
DAEMON=stunnel_thread
CONFIG=/etc/stunnel/stunnel.conf

test -x $PATH_DAEMON/$DAEMON || exit 0

case "$1" in
start)
$PATH_DAEMON/$DAEMON $CONFIG
;;
stop)
        pkill -f $DAEMON
;;
restart)
        pkill -f $DAEMON
        sleep 3
        $PATH_DAEMON/$DAEMON $CONFIG
;;
*)
        echo "Usage: $0 {start|stop|restart}"
        exit 1
;;
esac
exit 0

Добавляем в автозагрузку

sudo update-rc.d stunnel_thread defaults

6. Troubleshooting

Частые ошибки, с которыми я столкнулся и можете столкнуться вы.
1) Не открывается сайт через stunnel
Что нужно сделать:
— проверить адрес и порт в конфиге
— в браузере открывать не только сам адрес stunnel, а полный путь.
Пример: не 10.0.0.123:8181, а 10.0.0.123:8181/se/login
2) Следующие ошибки в логе

Error 0x40 ((unknown)) returned by CertVerifyCertificateChainPolicy!

и/или

Error 0x40 when validate certificate

Ошибка возникает из-за того, что stunnel не может проверить сертификат по списку отзыва CRL. Проблему можно решить тремя вариантами (но с разной эффективностью):
— периодически вручную импортировать актуальный CRL с помощью certmgr;
— отключить проверку сертификатов в TLS в файле настроек CSP

/etc/opt/cprocsp/config.ini

— в файле настроек CSP (написан выше) указать корректный путь к библиотеке curl.

"libcurl.so" = "/usr/lib/i386-linux-gnu/libcurl.so.4"

7. Ссылки

Тестовый Удостоверяющий Центр КриптоПро
КриптоПро CSP
КриптоПро Stunnel
Официальные мануалы по КриптоПро Stunnel

Ключевые слова для поисковых систем: настройка БКИ, автоматизированый обмен с бюро кредитных историй, автоматическое получение отчетов с бюро кредитных историй, КриптоПро и БКИ, обмен с бюро кредитных историй, шифрование трафика криптопро, криптопро на linux

ХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХ
Andrey, [05.10.18 11:28]
[В ответ на Дмитрий Давидович]
У них кривой сертификат сервера.
Диагностировать можно так:
[root@test-x64-centos7 ~]# /opt/cprocsp/bin/amd64/csptestf -tlsc -server zakupki.gov.ru -v -nosave -savecert /tmp/t.p7b
[root@test-x64-centos7 ~]# /opt/cprocsp/bin/amd64/certmgr -list -file /tmp/t.p7b
=============================================================================
1——-
Issuer              : E=uc_fk@roskazna.ru, S=г. Москва, INN=007710568760, OGRN=1047797019830, STREET=»улица Ильинка, дом 7″, L=Москва, C=RU, O=Федеральное казначейство, CN=Федеральное казначейство
Subject             : INN=007710568760, OGRN=1047797019830, STREET=»ул. Проспект Мира, 105″, E=777@roskazna.ru, C=RU, S=г. Москва, L=Москва, O=Федеральное казначейство, CN=zakupki.gov.ru
Serial              : 0x6F064FA71C24CD7E2CE6FAAEA927D8C7EC69A35F
SHA1 Hash           : c369b560ce239beddb2fc12b4884dee1cfc923aa
SubjKeyID           : 10c6d12e7cd886d022bcdfea4cbe10e32acf82bc
Signature Algorithm : ГОСТ Р 34.11/34.10-2001
PublicKey Algorithm : ГОСТ Р 34.10-2001 (512 bits)
Public key          : 04 40 1b ce 0b 15 6a 4a 57 e0 1b d8 39 ee 86 83
                      32 70 ea db fd f5 39 c1 e6 de 3c c8 be 10 81 03
                      5d da 0b 3b 09 7a d3 0d 46 33 58 11 3b 20 94 99
                      fe 04 fe 8e e6 bc 32 53 ff 2e 71 10 8e e2 12 a1
                      52 cf
Not valid before    : 21/12/2017  06:06:33 UTC
Not valid after     : 21/03/2019  06:06:33 UTC
PrivateKey Link     : No
Subject Alt Names
  UPN               : ▒㌴ (2.5.4.12)
  URL               : 0
CDP                 : http://crl.roskazna.ru/crl/ucfk.crl
CDP                 : http://crl.fsfk.local/crl/ucfk.crl
Extended Key Usage  : 1.3.6.1.5.5.7.3.1
…

При наличии расширения Subject Alt Names RFC требует сверять имя сервера с ним, а не с полем CN. Поэтому CN=zakupki.gov.ru отброшен, а в Subject Alt Names хрень.

ХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХ
Andrey, [05.10.18 11:35]
Но мы сделали параметр для отключения требований RFC6125, который может подвергнуть систему опасности для атак с подменой сертификатов:
[root@test-x64-centos7 ~]# /opt/cprocsp/sbin/amd64/cpconfig -ini ‘configparameters’ -add long Rfc6125_NotStrict_ServerName_Check 1
[root@test-x64-centos7 ~]# /opt/cprocsp/bin/amd64/csptestf -tlsc -server zakupki.gov.ru -nosave
HDEContext expired: OK if file is completely downloaded
Reply status: HTTP/1.1 200 OK
1 connections, 586 bytes in 0.186 seconds;
Total: SYS: 0,020 sec USR: 0,140 sec UTC: 0,230 sec
[ErrorCode: 0x00000000]

#  /opt/cprocsp/bin/amd64/csptest -keyset -verifycontext
CSP (Type:80) v4.0.9017 KC2 Release Ver:4.0.9944 OS:Linux CPU:AMD64 FastCode:READY:AVX.
AcquireContext: OK. HCRYPTPROV: 25072019
GetProvParam(PP_NAME): Crypto-Pro GOST R 34.10-2012 KC2 CSP
Total: SYS: 0,000 sec USR: 0,000 sec UTC: 0,010 sec
[ErrorCode: 0x00000000]

L K, [15.12.18 14:12]
[В ответ на Andrey]
root@astra:~# /opt/cprocsp/sbin/amd64/cpconfig -ini ‘configparameters’ -add long Rfc6125_NotStrict_ServerName_Check 1
root@astra:~#  /opt/cprocsp/bin/amd64/csptestf -tlsc -server zakupki.gov.ru -nosave
HSECPKG_ATTR_PACKAGE_INFO not supported.
Error 0x800b010f (CERT_E_CN_NO_MATCH) returned by CertVerifyCertificateChainPolicy!
An error occurred in running the program.
/dailybuildsbranches/CSP_4_0/CSPbuild/CSP/samples/csptest/WebClient.c:764:Error authenticating server credentials!
Error number 0x800b010f (2148204815).
CN-имя сертификата не совпадает с полученным значением.
Total: SYS: 0,070 sec USR: 0,290 sec UTC: 1,720 sec
[ErrorCode: 0x800b010f]

#  /opt/cprocsp/bin/amd64/csptest -keyset -verifycontext
CSP (Type:80) v4.0.9017 KC2 Release Ver:4.0.9944 OS:Linux CPU:AMD64 FastCode:READY:AVX.
AcquireContext: OK. HCRYPTPROV: 25072019
GetProvParam(PP_NAME): Crypto-Pro GOST R 34.10-2012 KC2 CSP
Total: SYS: 0,000 sec USR: 0,000 sec UTC: 0,010 sec
[ErrorCode: 0x00000000]

Какой релиз КриптоПро нужно ставить?

ХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХ 
Andrey, [12.12.18 18:12]
Вышел финальный релиз КриптоПро CSP 4.0 R4 (сборка 4.0.9963 Abel):
https://www.cryptopro.ru/forum2/default.aspx?g=posts&m=97735#post97735
На дистрибутив получено положительное заключение, скоро будет сертификат.
В тестовых целях её можно установить из нашего репозитория:
https://www.cryptopro.ru/forum2/default.aspx?g=posts&t=12507

ХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХ
Andrey, [15.12.18 23:07]
[В ответ на L K]
В любом случае лучше ставить последний, на который сейчас получаем сертификат — 4.0.9963 Abel. В 9944 в этой команде должна помочь замена long на bool и 1 на true.

ХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХХ

Проблема решилась установкой КриптоПро версии 4.0.9963 и действиями как написал Андрей, за что ему огромное спасибо.

user@astra:~$  /opt/cprocsp/bin/amd64/csptest -keyset -verifycontext
CSP (Type:80) v4.0.9019 KC2 Release Ver:4.0.9963 OS:Linux CPU:AMD64 FastCode:READY:AVX.
AcquireContext: OK. HCRYPTPROV: 26834115
GetProvParam(PP_NAME): Crypto-Pro GOST R 34.10-2012 KC2 CSP
Total: SYS: 0,000 sec USR: 0,000 sec UTC: 0,020 sec
[ErrorCode: 0x00000000]

user@astra:~$ sudo  /opt/cprocsp/sbin/amd64/cpconfig -ini ‘configparameters’ -add long Rfc6125_NotStrict_ServerName_Check 1

user@astra:~$ sudo  /opt/cprocsp/bin/amd64/csptestf -tlsc -server zakupki.gov.ru -nosave
HSECPKG_ATTR_PACKAGE_INFO not supported.
DContext expired: OK if file is completely downloaded
Reply status: HTTP/1.1 200 OK
1 connections, 589 bytes in 1.221 seconds;
Total: SYS: 0,040 sec USR: 0,320 sec UTC: 1,300 sec
[ErrorCode: 0x00000000]

Для  4.0.9944
user@astra:~$ /opt/cprocsp/bin/amd64/csptest -keyset -verifycontext
CSP (Type:80) v4.0.9017 KC2 Release Ver:4.0.9944 OS:Linux CPU:AMD64 FastCode:READY:AVX.
AcquireContext: OK. HCRYPTPROV: 20066515
GetProvParam(PP_NAME): Crypto-Pro GOST R 34.10-2012 KC2 CSP
Total: SYS: 0,000 sec USR: 0,000 sec UTC: 0,030 sec
[ErrorCode: 0x00000000]
user@astra:~$ sudo  /opt/cprocsp/sbin/amd64/cpconfig -ini ‘configparameters’ -add bool Rfc6125_NotStrict_ServerName_Check true
user@astra:~$ sudo  /opt/cprocsp/bin/amd64/csptestf -tlsc -server zakupki.gov.ru -nosave
HSECPKG_ATTR_PACKAGE_INFO not supported.
DContext expired: OK if file is completely downloaded
Reply status: HTTP/1.1 200 OK
1 connections, 589 bytes in 0.388 seconds;
Total: SYS: 0,050 sec USR: 0,200 sec UTC: 0,450 sec
[ErrorCode: 0x00000000]