For IBM AD V5.1.0.7 and earlier versions, the communication
between IBM AD Validation Server and IBM AD Connect for Mainframe is unencrypted socket session. Beginning with version
5.1.0.8, the optional secure communication, which uses the TLS protocol, is supported by using the
AT-TLS feature of IBM
z/OS® Communication Server.
The TLS protocol is a client or server cryptographic protocol. It is based on the earlier Secure
Sockets Layer (SSL) specifications that are developed by Netscape Corporation for securing
communications that use Transmission Control Protocol/Internet Protocol (TCP/IP) sockets. The TLS
and SSL protocols are designed to run at the application level. Therefore, typically, an application
must be designed and coded to use TLS/SSL protection. On z/OS, the System SSL component of the Cryptographic Services element implements the full suite
of SSL and TLS protocols (SSL V2, SSL V3, TLS V1.0, TLS V1.1, and TLS V1.2 as of this writing),
including a robust set of application programming interfaces (APIs) for z/OS C and C++ applications to use.
To make the TLS or SSL protocol more accessible to z/OS
applications, z/OS Communications Server V1R7 introduced the
AT-TLS feature. AT-TLS starts TLS or SSL primitives in the TCP layer of the TCP/IP stack on behalf
of application programs, based on policy files that describe the application traffic and how to
protect it. With AT-TLS, z/OS applications that are written
in almost any language can enjoy full TLS or SSL protection without requiring source code changes.
AT-TLS allows socket applications to access encrypted sessions by invoking system SSL within the
transport layer of the TCP/IP stack.
AT-TLS policy is read, parsed, and installed into the TCP/IP stack by the z/OS Communication Server Policy Agent (PAGENT), which implements policy-based
networking for the z/OS environment. The application
continues to send and receive clear text over the socket, but data that is sent over the network is
protected by the System SSL component. For more information about policy-based networking, see z/OS Communications
Server V2R3: IP Configuration Guide, SC27-3650-30.
To configure IBM AD Connect for Mainframe as a server with
TLS/SSL support, perform the following steps. If you have already configured AT-TLS for other
servers on your z/OS system, begin at step 5.
PAGENT runs as a UNIX process, so it
can be started either from the UNIX System Service shell or
as a z/OS started task. The following example shows how to
start PAGENT as a z/OS started task. You can also find this
sample file in TCPIP.SEZAINST(EZAPAGSP).Figure 1. Example: PAGENT JCL started
procedure
To
run with the required configuration, you can use environment variables that are configured either in
an IBM Multiple Virtual Storage (IBM
MVS™) data set, or in a z/OS
UNIX file that is specified by the STDENV data definition
(DD). The following example shows how to configure the environment variables for PAGENT. You can
also find this sample file in SYS1.TCPPARMS(PAGENTV).Figure 2. Example:
SYS1.TCPPARMS(PAGENTV) data set containing PAGENT environment
variables
TZ=PST8PDT7
PAGENT_CONFIG_FILE=//'SYS1.TCPPARMS(PAGENTP)' == > in unix default file is /etc/pagent.conf
PAGENT_LOG_FILE=SYSLOGD == > it could be /tmp/pagent.log
TZ
Specifies the local time zone for the PAGENT process. In the preceding example, it is set to
Pacific Standard Time (GMT -8) and Pacific Daylight Saving Time (GMT -7).
PAGENT_CONFIG_FILE
Specifies the PAGENT configuration file. In the preceding example, the PAGENT configuration file
is a member of the SYS1.TCPPARMS data set. For more information about PAGENT configuration files,
see step 3.
PAGENT_LOG_FILE
Specifies the name of the log file that is used by PAGENT. In the preceding example, it is set
to log PAGENT messages into file SYSLOGD.
Before you start the PAGENT started task, make sure that the appropriate security
authorizations have been set. See the following step for details.
Define the security authorization for PAGENT.
The policies that are managed by PAGENT can affect system operation significantly. Therefore,
you must restrict the access of z/OS user IDs to PAGENT. To
do this, define certain resources and controls in the security manager product of the system, for
example, RACF. To set up the security definitions of PAGENT
to RACF, follow the steps:
Define the PAGENT started task to RACF.
To set up the
PAGENT started task to RACF, you must define a profile to the
STARTED RACF generic resource class by using the
RDEFINE command. The following example shows the RACF commands that are used to set up the PAGENT started task. You can also find
this sample file in SIAYSAMP(RACFPAG1).Figure 3. Example: RACF commands to define the PAGENT started task to RACF
If
you also want to log messages into the SYSLOGD file that is mentioned in Figure 2, use the
RDEFINE command to define a profile for SYSLOGD to the STARTED class, for
example, RDEFINE STARTED SYSLOGD.*.
Define the PAGENT user ID.
Before z/OS V1R13, the user
ID for the PAGENT started task must have the z/OS
UNIX superuser authority, which means the z/OS
UNIX user ID (UID) for this user must be set to 0, and you
must assign a default group (DFLTGRP) for the user ID. In the following example, PAGENT, the policy
agent, runs under the z/OS user ID PAGENT. The RACF command is used to define the PAGENT user ID to the OMVSGRP
default group with an OMVS segment that has a UID of 0. You can also find this sample file in
SIAYSAMP(RACFPAG2).Figure 4. Example: RACF command to define a user ID for the PAGENT started
task
If
the SHARED.IDS profile has been defined in the UNIXPRV class, the UID value must be unique. If you
want to use a UID value that is already in use, you can indicate that you intend to share the UID
value across z/OS user IDs by adding the SHARED keyword to
the UID parameter of the ADDUSER command.
If you also want to log messages
into the SYSLOGD file that is mentioned in Figure 2, use the
ADDUSER command to define a user ID with the superuser authority for the SYSLOGD
started task, for example, ADDUSER SYSLOGD DFLTGRP(OMVSGRP) OMVS(UID(0)
HOME('/')).
Associate the PAGENT user ID with the PAGENT started task.
Use the RACF
RALTER command to associate the PAGENT user ID, which is created from the
preceding step, to the PAGENT started task. See the following example. You can also find this sample
file in SIAYSAMP(RACFPAG3).Figure 5. Example: RACF command to associate the PAGENT user ID with the PAGENT started
task
If
you also want to log messages into the SYSLOGD file that is mentioned in Figure 2, use the
RALTER command to associate the SYSLOGD user ID with the SYSLOGD started task,
for example, RALTER STARTED PAGENT.* STDATA(USER(SYSLOGD)).
Give authorized users access to manage the PAGENT started task.
To restrict management access
to the PAGENT started task, you must define a profile that is named MVS.SERVMGR.PAGENT in the OPERCMDS RACF resource class,
and give only authorized users access to this facility. The following example shows the RACF commands that are used to configure this access. You can
also find this sample file in SIAYSAMP(RACFPAG4).Figure 6. Example: RACF commands to give authorized users access to manage the
PAGENT started
task
The pasearch
UNIX command can be used to display policy definitions. The
output from this command indicates whether policy rules are active and shows the parsed results of
the policy definition attributes. PAGENT, the policy agent is designed to ignore unknown attributes,
so misspelled attributes result in the default values being used.
The
pasearch output can be used to verify whether policies are defined correctly.
However, you might not want every user to be able to see the policy definitions. To restrict
unauthorized access to the pasearch command, a resource is defined to the
SERVAUTH RACF resource class. This type of resource can be
defined for each TCP/IP stack (TcpImage) and policy type (ptype) by using the following command:
EZB.PAGENT.sysname.TcpImage.ptype
sysname
The z/OS system name.
TcpImage
The procedure name of the TCP/IP stack that policy information is to be restricted to.
ptype
The policy type that is being requested. The following types are possible:
QoS
Policy Quality of Service
IDS
Policy Intrusion Detection System
IPSec
Policy Internet Protocol Security
TTLS
Policy AT-TLS
The following example shows the RACF commands that are
used to restrict access to the pasearch
UNIX command. In this example, only the USRT001 and USRT002
z/OS user IDs are permitted to use the
pasearch command for the TCPIP TCP/IP stack on the UTEC224 z/OS system. You can also find this sample file in SIAYSAMP(RACFPAG4).Figure 7. Example: RACF commands to
restrict unauthorized users access to the pasearch
UNIX
command
PAGENT is responsible for reading policies from configuration members in MVS data sets or in z/OS
UNIX file system files. Before you can define AT-TLS
policies, you must configure certain operational characteristics of PAGENT in a main configuration
file. The following statements can be contained in the main configuration file:
The TcpImage statement
The LogLevel statement
Define the TcpImage statements and appropriate logging level in the main configuration file.
The following example shows definitions in the main configuration file for a single TCP/IP stack
environment. You can also find this sample file in UNIX
default file /etc/pagent.conf or in SIAYSAMP(CNFGPAG). In this example, the
AT-TLS policies are defined in a separate image-specific configuration file, TTLSPOL, for the TCPIP
TCP/IP image.Figure 8. Example: PAGENT main configuration
file
# LogLevel statement
# SYSERR, OBJERR, PROTERR, and WARNING messages are logged.
LogLevel 15
#
# TcpImage statement
# TCP/IP image: TCPIP
# Path to image-specific configuration file: SYS1.TCPPARMS(TTLSPOL)
# FLUSH parameter specified to delete existing policy data in the
# stack on PAGENT start-up or when the configuration files change.
# PURGE parameter specified to delete active policy data from the
# stack and Policy Agent when PAGENT is shut down normally.
TcpImage TCPIP //'SYS1.TCPPARMS(TTLSPOL)' FLUSH PURGE
# Here is an alternate way to specify a common AT-TLS policy for the
# TCPIP image:
#
# TcpImage TCPIP FLUSH PURGE
# TTLSConfig //'SYS1.TCPPARMS(TTLSPOL)' FLUSH PURGE
#
As
mentioned in Figure 2,
/etc/pagent.conf is copied to SYS1.TCPPARMS(PAGENTP), and this configuration
file is specified by using the PAGENT_CONFIG_FILE environment variable in
SYS1.TCPPARMS(PAGENTV).
Configure AT-TLS.
AT-TLS support is enabled by specifying the TTLS parameter on the
TCPCONFIG statement in the TCP/IP profile data set. The information that is required to negotiate
secure connections is provided to the TCP/IP stack by the AT-TLS policies that are read, parsed, and
installed by PAGENT.
When AT-TLS is enabled and a new TCP connection is established, the AT-TLS
component in the TCP layer of the stack searches for an AT-TLS rule in the policy that matches the
characteristics, including local and remote IP addresses, local and remote ports, the connection
direction, and other characteristics. If such a rule is found, TLS protection is applied to the rule
according to the details that are specified in the AT-TLS action that is associated with the rule.
If no such a rule is found, the connection is not protected with TLS.
Tip: To enable
TTLS without modifying PROFILE.TCPIP, and without stopping and starting the TCP/IP stack, define a
separate file that contains the TCPCONFIG TTLS statement, and run the VARY TCPIP
OBEYFILE command, as shown in the following
example:
1. OBEYFILE called TTLSON in the SYS1.TCPPARMS data set:
TCPCONFIG TTLS
2. On the MVS console, issue this command:
VARY TCPIP,,O,DSN=SYS1.TCPPARMS(TTLSON)
To disable TTLS, perform the following actions:
1. OBEYFILE called TTLSOFF in the SYS1.TCPPARMS data set:
TCPCONFIG NOTTLS
2. On the MVS console, issue this command:
VARY TCPIP,,O,DSN=SYS1.TCPPARMS(TTLSOFF)
To
make permanent changes for TLS support, update the TCP/IP configuration in PARMLIB:
Add the TTLS parameter to the TCPCONFIG section.
Add the PAGENT parameter to the AUTOLOG section.
Optionally, you can set up TTLS stack initialization access control for AT-TLS. When
AT-TLS is started during TCP/IP stack initialization, there might be a delay between the time that
the stack comes up and when PAGENT successfully installs policy information into the stack. This
situation can leave a window of time where the connections that are intended to be protected by
AT-TLS can be established without that protection. While this window is open, to prevent such
connections from being established, define a profile for the EZB.INITSTACK resource in the SERVAUTH
RACF resource class. This type of resource can be defined for
each TCP/IP stack by using the following command:
EZB.INITSTACK.sysname.TcpImage
sysname
The system name that is assigned to the z/OS LPAR.
TcpImage
The procedure name of the TCP/IP stack to which access is to be controlled.
The following example shows how to set up TCP/IP stack initialization access control for AT-TLS
in the profile. You can also find this sample file in SIAYSAMP(STKINIT).Figure 9. Example: RACF commands to set up TCP/IP stack
initialization access control for
AT-TLS
With
such a profile in place, you can control which applications are allowed to establish TCP connections
before PAGENT is started. To do so, provide the READ access to the
EZB.INITSTACK.sysname.TcpImage profile in the SERVAUTH class
for the z/OS user IDs under which such applications will run.
Before connections are enabled, all other applications are forced to wait until PAGENT is
initialized and its policies are installed into the stack. For more information, see z/OS Communications
Server V2R3: IP Configuration Guide, SC27-3650-30.
As part of the TLS configuration process in IBM AD Validation Server, a personal certificate to present IBM AD Validation Server is created. This personal certificate must be signed with a certification authority (CA) certificate or a self-signed certificate. Save a copy of this CA certificate or self-signed certificate on the mainframe.
Ensure that the certificate file is downloaded from IBM AD Validation Server in the correct format. The example in the configuration instructions for IBM AD Validation Server creates a DER encoded X.509 certificate that is in the Base64
format, so download this certificate in the ASCII mode if using FTP. An alternative to FTP is to
copy and paste the certificate contents. The mainframe data set that contains the certificate must
be VB with a LRECL of at least 256 bytes.
Important: The personal certificate that is presented to IBM AD Connect for Mainframe on behalf of the IBM AD Validation Server must have the Common Name field equal to the numeric IP address of the Windows machine where the IBM AD Validation Server is running. Do not use the host name. Confirm this with the person who performs TLS configurations in IBM AD Validation Server.
Import the personal certificated that is saved in the preceding step to the RACF database as TRUSTED. Create an RACF key ring in the RACF database for the user ID that the skeleton in ChangeMan will run with. Connect the imported personal certificate to the key ring.
The following example shows how to import a CA certificate to the RACF database, create a key ring, and connect a user ID to the key ring. The user ID needs the read access to IRR.DIGTCERT.LISTRING in the FACILTY class. You can also find this sample file in IAYV510.SIAYSAMP(VCERTJCL) or IAYV510.SAMPLIB(VCERTJCL).Figure 10. Example: Sample JCL to import a CA certificate to the RACF database, create a key ring, and connect a user ID to the key ring
//jobname JOB ’SAMPLE JCL’, MSGLEVEL=1,MSGCLASS=H,CLASS=A,REGION=0M
//****************************************************************************************************
//* JOB INSTRUCTIONS: *
//* ENTER VALID JOBCARD ABOVE. *
//*1. CHANGE ++CERTIFICATE-DATASET++ TO THE SEQUENTIAL FILE CONTAINING
//* CA CERT OR SELF SIGNED CERTIFICATE.
//* 2. CHANGE ++LABEL ++ TO APPROPRIATE LABEL NAME .
//* 3. CHANGE ++RINGNAME++ TO YOUR KEY RING NAME. EX: VALRING *
//* 4. CHANGE ++USERID++ TO THE USERID SUBMITTING THE
//* COMPILE SKELETON(S) IN CHANGEMAN
//****************************************************************************************************
//* AT THE END OF THIS JOB YOU SHOULD HAVE THE FOLLOWING: *
//* KEYRING : ++RINGNAME++ *
//* CA CERT/SELF SIGNED CERT CONNECTED TO THE RING OWND BY USERID
//****************************************************************************************************
//* * ADD CA CERT/SELF-SIGNED CERTIFICATE TO THE RACF DATABASE
//* * CERTAUTH - CERTIFICATE AUTHORITY TYPE.
//****************************************************************************************************
//****************************************************************************************************
//STEP1 EXEC PGM=IKJEFT01,REGION=512K
//SYSPRINT DD SYSOUT=*
//SYSOUT DD SYSOUT=*
//SYSTSPRT DD SYSOUT=*
//SYSTSIN DD *
RACDCERT CERTAUTH ADD('++CERTIFICATE DATASET++') +
WITHLABEL('++LABEL++’) TRUST
SETROPTS RACLIST(DIGTCERT DIGTRING) REFRESH
SETROPTS RACLIST(FACILITY) REFRESH
/*
//*********************************************************************************************
//*Create RING and Connect CA cert to it.***********************************************
//*RING to be owned by USERID submitting Changeman skeleton Job.**************
//*********************************************************************************************
//STEP2 EXEC PGM=IKJEFT01,REGION=512K
//SYSPRINT DD SYSOUT=*
//SYSOUT DD SYSOUT=*
//SYSTSPRT DD SYSOUT=*
//SYSTSIN DD *
RACDCERT ID(++USERID++) ADDRING(++RINGNAME++)
RACDCERT ID(++USERID++) CONNECT(CERTAUTH LABEL('++LABEL++') +
RING(++RINGNAME++) +
default
SETROPTS RACLIST(DIGTCERT DIGTRING) REFRESH
SETROPTS RACLIST(FACILITY) REFRESH
/*
//**************************************************************************************************
//*PERMIT USERID to access the RING
//*REFRESH CLASSES.
//**************************************************************************************************
//PERMIT EXEC PGM=IKJEFT01,REGION=512K
//SYSPRINT DD SYSOUT=*
//SYSOUT DD SYSOUT=*
//SYSTSPRT DD SYSOUT=*
//SYSTSIN DD *
PERMIT IRR.DIGTCERT.LISTRING CLASS(FACILITY) ID(++USERID++) ACCESS(READ)
SETR RACLIST (FACILITY) REFRESH
/*
Set up the PAGENT policy for the port that is used by IBM AD Validation Server.
Because IBM AD Validation Server runs on the Windows machine, and the batch job runs on the mainframe as the client, the PAGENT policy for the secured port needs to be configured for the outbound direction. The following example shows how to set up the PAGENT policy. You can also find this sample file in IAYV510.SIAYSAMP(VALTTLS).Figure 11. Example:
PAGENT policy
setup
TTLSRule VAL_SERVER
{
RemotePortRange ++port-number++
Direction Outbound
TTLSGroupActionRef grp_Enabled
TTLSEnvironmentActionRef AD_VALIDATE
}
##
}
##
TTLSGroupAction grp_Enabled
{
TTLSEnabled On
Trace 255 # Errors to syslogd & IP joblog
#Trace 254 # Everything to syslogd
}
##----------------------------
TTLSEnvironmentAction AD_VALIDATE
{
HandshakeRole Client
TTLSKeyRingParms
{
Keyring ++RING-NAME++ # Keyring must be owned by the userid with which Changeman skeleton
Job will run
}
TTLSCipherParmsRef TLS_12 # Ciphers modern acceptable
##
TTLSEnvironmentAdvancedParms
{
ApplicationControlled Off
## TSLV1.2 only on z/OS 2.1
TLSV1.2 On # latest tls only
TLSV1.1 Off # TLSv1 & TLSv1.1 are on by default
TLSV1 Off # TLSv1 & TLSv1.1 are on by default
SSLV3 Off #
SSLV2 Off #
}
}
##
#
TTLSCipherParms TLS_12
{
# RSA, DSS auth with AES and SHA-2
V3CipherSuites TLS_DHE_RSA_WITH_AES_256_CBC_SHA256
V3CipherSuites TLS_DHE_DSS_WITH_AES_256_CBC_SHA256
V3CipherSuites TLS_DH_RSA_WITH_AES_256_CBC_SHA256
V3CipherSuites TLS_DH_DSS_WITH_AES_256_CBC_SHA256
V3CipherSuites TLS_DHE_RSA_WITH_AES_128_CBC_SHA256
V3CipherSuites TLS_DHE_DSS_WITH_AES_128_CBC_SHA256
V3CipherSuites TLS_DH_RSA_WITH_AES_128_CBC_SHA256
V3CipherSuites TLS_DH_DSS_WITH_AES_128_CBC_SHA256
V3CipherSuites TLS_RSA_WITH_AES_256_CBC_SHA256
V3CipherSuites TLS_RSA_WITH_AES_128_CBC_SHA256
# RSA, DSS auth with AES and SHA-1
V3CipherSuites TLS_DHE_RSA_WITH_AES_256_CBC_SHA
V3CipherSuites TLS_DHE_DSS_WITH_AES_256_CBC_SHA
V3CipherSuites TLS_DH_RSA_WITH_AES_256_CBC_SHA
V3CipherSuites TLS_DH_DSS_WITH_AES_256_CBC_SHA
V3CipherSuites TLS_RSA_WITH_AES_256_CBC_SHA
V3CipherSuites TLS_DHE_RSA_WITH_AES_128_CBC_SHA
V3CipherSuites TLS_DHE_DSS_WITH_AES_128_CBC_SHA
V3CipherSuites TLS_DH_RSA_WITH_AES_128_CBC_SHA
V3CipherSuites TLS_DH_DSS_WITH_AES_128_CBC_SHA
V3CipherSuites TLS_RSA_WITH_AES_128_CBC_SHA
# RSA, DSS auth with 3DES and SHA-1
V3CipherSuites TLS_DHE_RSA_WITH_3DES_EDE_CBC_SHA
V3CipherSuites TLS_DHE_DSS_WITH_3DES_EDE_CBC_SHA
V3CipherSuites TLS_DH_RSA_WITH_3DES_EDE_CBC_SHA
V3CipherSuites TLS_DH_DSS_WITH_3DES_EDE_CBC_SHA
V3CipherSuites TLS_RSA_WITH_3DES_EDE_CBC_SHA
}
After you define the AT-TLS rules, refresh or update the policy agent for
the AT-TLS rules to take effect.
If PAGENT is started before you define the AT-TLS rules, you must refresh or update PAGENT by
using the MVS MODIFY command:
F PAGENT,REFRESH
The REFRESH command triggers PAGENT to reread and process the policy files.
If the FLUSH parameter is specified on the TcpImage or discipline configuration
statement, the REFRESH command triggers the FLUSH processing.
Because the FLUSH processing deletes and re-installs all policies, you must use this command
only if you suspect that policies have become out of sync between the TCP/IP stack and PAGENT. Note
that one consequence of triggering the FLUSH processing is that policy statistics that are being
collected in the TCP/IP stack are reset.
F PAGENT,UPDATE
The UPDATE command triggers PAGENT to reread and process the policy files.
This command differs slightly from the REFRESH command. By using the
UPDATE command, PAGENT installs and removes only the new, changed, or deleted
policies from the stack. Unchanged policies are unaffected. Therefore, it is suggested to use the
UPDATE command in most cases.