Re: How should I send file to another sysplex securely.

2021-07-24 Thread Radoslaw Skorupka

There are many ways to skin a cat.
You can rely on industry standards like certificates and CA's.
You can also use your own methods.
You can use both. Yes, double security.

What I could do:
1. Establish some secure file transfer. It can be FTPS, sftp, commercial 
MFT applications like Sterling Connect, MQ MFT, etc.
2. Just to sleep better I would also encipher the dataset before it is 
sent. Again, it can be done using commercial tools like Encryption 
Facility or Megacryption or other. Of course it is feasible to encrypt 
dataset using your own tools.


Why two methods?
Well, both are safe ...until someone made a mistake. Assuming some 
separation of duties it would be more idiotproof and proof of malicious 
insider.


--
Radoslaw Skorupka
Lodz, Poland




W dniu 22.07.2021 o 16:07, Colin Paice pisze:

I was wondering the best way customers send sensitive data between z/OS
images.
I was thinking about exporting one's private certificates.

1. I can create a dataset of the private certificates on system 1 and
have it encrypted.   I can send it to the other system.   How can I decrypt
it on the remote system as it needs shared certificates?  It seems a
chicken and egg problem
2. I can put a password on the file through JCL and use FTPS to send
it.   This could easily be broken

This is hypothetical, but I would be interested in how to do it.

Colin Paice

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN


--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN


Re: How should I send file to another sysplex securely.

2021-07-23 Thread Grant Taylor

On 7/22/21 6:21 PM, Charles Mills wrote:

Agreed. By "roll your own" I was referring to


1)  Create an asymmetric public + private key pair on the destination
system.
2)  Transfer the destination system's public key to the source system.
3)  Create a symmetric key on the source system.


Etc.


If that's "roll your own" cryptography, then how do you use Pervasive 
Encryption without "rolling your own" cryptography?


You must have keying material, and I'd be flabbergasted if IBM shipped 
said keying material with the system.


You have to transfer that keying material between systems you want to 
communicate.


How do you avoid "rolling your own" cryptography when you make the 
investment and create the aforementioned SFTP server?




--
Grant. . . .
unix || die

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN


Re: How should I send file to another sysplex securely.

2021-07-23 Thread Grant Taylor

On 7/22/21 6:17 PM, Mike Schwab wrote:
Since a lot of chips a manufactured in China, a device could be 
sending ... your data,


Theoretically yes.  I'm not going to speculate on the probability that 
such is happening.  Though Hanlon's Razor comes to mind.


But for it to be sending your data it would have to ... actually send 
the data.  And in doing so would present symptoms of sending data. 
These symptoms can be detected.  As such it's quite possible to detect 
if this is happening.


I say possible, because not everybody is looking for such data.



--
Grant. . . .
unix || die

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN


Re: How should I send file to another sysplex securely.

2021-07-23 Thread Grant Taylor

On 7/22/21 6:09 PM, Charles Mills wrote:

Guys, this is the problem with inventing your own solution.


You didn't elucidate what the (or a) problem is.


Public keys are, well, public.


Yes, that's the very nature of a /public/ key.

The new fashion in fact is to NOT trust internal networks. You really 
don't know how far an intruder may have intruded, so you should assume 
every communication channel is insecure.

*HEAVYsigh*

Yes.  Zero Trust *is* /a/ thing.  But I think it's *MUCH* /more/ of a 
buzz word trying to sell snake oil than it is a real problem.


As Bruce Schneier is famous for saying "Trust the Math".

If I compare (a cryptographic hash of) the public key on SYS1 and SYS2 
and they are the same, combined with the fact that SYS1's corresponding 
key can decrypt what SYS2 encrypts with it's copy of SYS1's public key, 
then I think the math dictates that SYS1 and SYS2 are actually talking 
to each other, no matter how (in)secure the intervening network(s) are.


And if a malicious actor can intercept data flowing across a HyperSocket 
between two LPARs and perform a full in the middle attack, ... you've 
got MUCH bigger problems.


In the new era of cloud and partners and bring your 
own device, exactly what is an internal and what is 
an external network?


The two CECs sitting next to each other in the same room with cables 
running directly between them seems to definitely qualify as "internal" 
to me.


You can "what if" yourself to death.  Or you can "trust the math". 
Verify (a hash of) the public key on the source and destination system. 
If the key is the same, you are quite likely safe to move data across 
the wire.  If someone can spoof the output from the commands to display 
(the hash of) the public key and your terminal as an active in the 
middle attack ... you have bigger problems.  I'm not even sure that IBM 
can help you.


This new paradigm is called "Zero 
Trust." https://csrc.nist.gov/publications/detail/sp/800-207/final. I 
have a presentation on Zero Trust coming up courtesy of New Era, 
but we don't have a date yet. A good month or so out.


I'm intrigued.  But obviously I'm a tad bit skeptical.



--
Grant. . . .
unix || die

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN


Re: How should I send file to another sysplex securely.

2021-07-22 Thread Charles Mills
Agreed. By "roll your own" I was referring to

>1)  Create an asymmetric public + private key pair on the destination 
>system.
>2)  Transfer the destination system's public key to the source system.
>3)  Create a symmetric key on the source system.

Etc.

Charles


-Original Message-
From: IBM Mainframe Discussion List [mailto:IBM-MAIN@LISTSERV.UA.EDU] On Behalf 
Of Grant Taylor
Sent: Thursday, July 22, 2021 4:58 PM
To: IBM-MAIN@LISTSERV.UA.EDU
Subject: Re: How should I send file to another sysplex securely.

On 7/22/21 2:58 PM, Charles Mills wrote:
> I would say in no event does the OP want to "roll his own" or "cobble 
> something together out of bits and pieces."

I think we have different ideas of what "roll your own" means.

Personally, I don't believe that running some standard commands (at 
least from a unix perspective), transferring two files, and running some 
closely related commands to be "rolling your own".

At least not any more than creating JCL is "roll your own".

> This problem is what FTP does for a living.

Agreed.

> An investment in secure FTP is an investment in the future, not just 
> this one problem.

Yes.

Though, sometimes such an /investment/ means a LOT more work, especially 
if something is going to be persistent and need to adhere to corporate 
security policies / scans / etc.

Often, especially for one off cases, doing a little bit more work 
manually is the simpler and faster of the solutions.

> Oh! In Step 3 below, add to the sentence "... using a secure 
> cryptographic-quality random-number generator."

Agreed.

I expect that any contemporary / patched operating system to 
sufficiently address this concern.  Especially when following vendor 
best practices regarding cryptographic utilities that they provide.

> Again, you don't want to roll your own on this. Wy too many traps 
> for the unwary.

See above.  There is a big difference in putting some commands in JCL vs 
coding your own cryptographic algorithms, including the math and keying 
algorithms.  There is an apt saying for that "friends don't let friends 
create their own cryptographic algorithms".  Friends help friends use 
well established cryptographic algorithms in the proper way.



-- 
Grant. . . .
unix || die

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN


Re: How should I send file to another sysplex securely.

2021-07-22 Thread Mike Schwab
Since a lot of chips a manufactured in China, a device could be
sending China your data, which is why Huawei is being banned from
communications networks.

On Thu, Jul 22, 2021 at 7:09 PM Charles Mills  wrote:
>
> Guys, this is the problem with inventing your own solution.
>
> Public keys are, well, public. You can transfer them over an insecure medium. 
> You can put them on your Web site. You can announce them over the PA. The TLS 
> protocol transfers public keys over the communication link before it is 
> trusted.
>
> > If the OP can't safely get a public [key] copied between LPARs / CECs in a 
> > trusted network
>
> The new fashion in fact is to NOT trust internal networks. You really don't 
> know how far an intruder may have intruded, so you should assume every 
> communication channel is insecure. In the new era of cloud and partners and 
> bring your own device, exactly what is an internal and what is an external 
> network? (This new paradigm is called "Zero Trust." 
> https://csrc.nist.gov/publications/detail/sp/800-207/final. I have a 
> presentation on Zero Trust coming up courtesy of New Era, but we don't have a 
> date yet. A good month or so out.)
>
> Charles
>
>
> -Original Message-
> From: IBM Mainframe Discussion List [mailto:IBM-MAIN@LISTSERV.UA.EDU] On 
> Behalf Of Grant Taylor
> Sent: Thursday, July 22, 2021 4:48 PM
> To: IBM-MAIN@LISTSERV.UA.EDU
> Subject: Re: How should I send file to another sysplex securely.
>
> On 7/22/21 3:17 PM, Paul Gilmartin wrote:
> > It lacks authentication and does not prevent MITM attacks:
>
> I think we might be talking about two slightly, but distinctly,
> different scenarios.
>
> I took the OP's statement to be talking about needing to move data from
> one LPAR / CEC on the left side of the room to another LPAR / CEC on the
> right side of the room.  Wherein the room and the network are trusted; a
> la. internal company network.
>
> What's more is I was anticipating the OP to be performing all of the
> steps.  As such, the OP could validate that the public key copied from
> the destination system to the source system was in fact one in the same.
>   Be it byte for byte comparison of hex output, or comparison of
> cryptographic hashes, or even IND$FILE transfers from the destination
> system to the source system via the common workstation / terminal emulator.
>
> Aside:  If the OP needs to do the transfer in conjunction with a fellow
> SYSOP from elsewhere in the world, they can get on the phone with each
> other (or use some other out of band communications method that they
> trust) to confirm public key.
>
> Further aside:  If the OP can't safely get a public copied between LPARs
> / CECs in a trusted network, then s/he has bigger problems.  If
> interlopers are messing with such a transfer, ... that's a *LOT* bigger
> problem.  Problems big enough that asking for help on a mailing list is
> quite likely not sufficient.
>
>
>
> --
> Grant. . . .
> unix || die
>
> --
> For IBM-MAIN subscribe / signoff / archive access instructions,
> send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN
>
> --
> For IBM-MAIN subscribe / signoff / archive access instructions,
> send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN



-- 
Mike A Schwab, Springfield IL USA
Where do Forest Rangers go to get away from it all?

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN


Re: How should I send file to another sysplex securely.

2021-07-22 Thread Charles Mills
Guys, this is the problem with inventing your own solution.

Public keys are, well, public. You can transfer them over an insecure medium. 
You can put them on your Web site. You can announce them over the PA. The TLS 
protocol transfers public keys over the communication link before it is trusted.

> If the OP can't safely get a public [key] copied between LPARs / CECs in a 
> trusted network

The new fashion in fact is to NOT trust internal networks. You really don't 
know how far an intruder may have intruded, so you should assume every 
communication channel is insecure. In the new era of cloud and partners and 
bring your own device, exactly what is an internal and what is an external 
network? (This new paradigm is called "Zero Trust." 
https://csrc.nist.gov/publications/detail/sp/800-207/final. I have a 
presentation on Zero Trust coming up courtesy of New Era, but we don't have a 
date yet. A good month or so out.)

Charles


-Original Message-
From: IBM Mainframe Discussion List [mailto:IBM-MAIN@LISTSERV.UA.EDU] On Behalf 
Of Grant Taylor
Sent: Thursday, July 22, 2021 4:48 PM
To: IBM-MAIN@LISTSERV.UA.EDU
Subject: Re: How should I send file to another sysplex securely.

On 7/22/21 3:17 PM, Paul Gilmartin wrote:
> It lacks authentication and does not prevent MITM attacks:

I think we might be talking about two slightly, but distinctly, 
different scenarios.

I took the OP's statement to be talking about needing to move data from 
one LPAR / CEC on the left side of the room to another LPAR / CEC on the 
right side of the room.  Wherein the room and the network are trusted; a 
la. internal company network.

What's more is I was anticipating the OP to be performing all of the 
steps.  As such, the OP could validate that the public key copied from 
the destination system to the source system was in fact one in the same. 
  Be it byte for byte comparison of hex output, or comparison of 
cryptographic hashes, or even IND$FILE transfers from the destination 
system to the source system via the common workstation / terminal emulator.

Aside:  If the OP needs to do the transfer in conjunction with a fellow 
SYSOP from elsewhere in the world, they can get on the phone with each 
other (or use some other out of band communications method that they 
trust) to confirm public key.

Further aside:  If the OP can't safely get a public copied between LPARs 
/ CECs in a trusted network, then s/he has bigger problems.  If 
interlopers are messing with such a transfer, ... that's a *LOT* bigger 
problem.  Problems big enough that asking for help on a mailing list is 
quite likely not sufficient.



-- 
Grant. . . .
unix || die

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN


Re: How should I send file to another sysplex securely.

2021-07-22 Thread Grant Taylor

On 7/22/21 5:42 PM, Lennie Dymoke-Bradshaw wrote:
There is a document by Philippe Richard of IBM France which documents 
this problem and demonstrates how to resolve it using a set of REXX 
routines written by Eysha Powers.


It is entitled "Transporting AES encrypted data keys from one z/OS host 
to another". As far as I can see it has no manual number.


This looks like it (at a quick glance).

https://www.ibm.com/support/pages/system/files/inline-files/Transporting_AES_DATA_keys_new.pdf

$ReadingList++



--
Grant. . . .
unix || die

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN


Re: How should I send file to another sysplex securely.

2021-07-22 Thread Grant Taylor

On 7/22/21 2:58 PM, Charles Mills wrote:
I would say in no event does the OP want to "roll his own" or "cobble 
something together out of bits and pieces."


I think we have different ideas of what "roll your own" means.

Personally, I don't believe that running some standard commands (at 
least from a unix perspective), transferring two files, and running some 
closely related commands to be "rolling your own".


At least not any more than creating JCL is "roll your own".


This problem is what FTP does for a living.


Agreed.

An investment in secure FTP is an investment in the future, not just 
this one problem.


Yes.

Though, sometimes such an /investment/ means a LOT more work, especially 
if something is going to be persistent and need to adhere to corporate 
security policies / scans / etc.


Often, especially for one off cases, doing a little bit more work 
manually is the simpler and faster of the solutions.


Oh! In Step 3 below, add to the sentence "... using a secure 
cryptographic-quality random-number generator."


Agreed.

I expect that any contemporary / patched operating system to 
sufficiently address this concern.  Especially when following vendor 
best practices regarding cryptographic utilities that they provide.


Again, you don't want to roll your own on this. Wy too many traps 
for the unwary.


See above.  There is a big difference in putting some commands in JCL vs 
coding your own cryptographic algorithms, including the math and keying 
algorithms.  There is an apt saying for that "friends don't let friends 
create their own cryptographic algorithms".  Friends help friends use 
well established cryptographic algorithms in the proper way.




--
Grant. . . .
unix || die

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN


Re: How should I send file to another sysplex securely.

2021-07-22 Thread Grant Taylor

On 7/22/21 3:17 PM, Paul Gilmartin wrote:

It lacks authentication and does not prevent MITM attacks:


I think we might be talking about two slightly, but distinctly, 
different scenarios.


I took the OP's statement to be talking about needing to move data from 
one LPAR / CEC on the left side of the room to another LPAR / CEC on the 
right side of the room.  Wherein the room and the network are trusted; a 
la. internal company network.


What's more is I was anticipating the OP to be performing all of the 
steps.  As such, the OP could validate that the public key copied from 
the destination system to the source system was in fact one in the same. 
 Be it byte for byte comparison of hex output, or comparison of 
cryptographic hashes, or even IND$FILE transfers from the destination 
system to the source system via the common workstation / terminal emulator.


Aside:  If the OP needs to do the transfer in conjunction with a fellow 
SYSOP from elsewhere in the world, they can get on the phone with each 
other (or use some other out of band communications method that they 
trust) to confirm public key.


Further aside:  If the OP can't safely get a public copied between LPARs 
/ CECs in a trusted network, then s/he has bigger problems.  If 
interlopers are messing with such a transfer, ... that's a *LOT* bigger 
problem.  Problems big enough that asking for help on a mailing list is 
quite likely not sufficient.




--
Grant. . . .
unix || die

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN


Re: How should I send file to another sysplex securely.

2021-07-22 Thread Lennie Dymoke-Bradshaw
Colin,

There is a document by Philippe Richard of IBM France which documents this 
problem and demonstrates how to resolve it using a set of REXX routines written 
by Eysha Powers. 

It is entitled "Transporting AES encrypted data keys from one z/OS host to 
another". As far as I can see it has no manual number. If you cannot find it 
from Philippe Richard, then I can send you a copy. The method makes use of 
standard ICSF calls to use EC keys to have the same AES data key installed into 
two distinct z/OS systems in a secure manner.

Once that key is installed in both systems, the data can be securely 
transferred in either direction.


Lennie Dymoke-Bradshaw
https://rsclweb.com 
‘Dance like no one is watching. Encrypt like everyone is.’


-Original Message-
From: IBM Mainframe Discussion List  On Behalf Of 
Colin Paice
Sent: 22 July 2021 15:08
To: IBM-MAIN@LISTSERV.UA.EDU
Subject: How should I send file to another sysplex securely.

I was wondering the best way customers send sensitive data between z/OS images.
I was thinking about exporting one's private certificates.

   1. I can create a dataset of the private certificates on system 1 and
   have it encrypted.   I can send it to the other system.   How can I decrypt
   it on the remote system as it needs shared certificates?  It seems a
   chicken and egg problem
   2. I can put a password on the file through JCL and use FTPS to send
   it.   This could easily be broken

This is hypothetical, but I would be interested in how to do it.

Colin Paice

--
For IBM-MAIN subscribe / signoff / archive access instructions, send email to 
lists...@listserv.ua.edu with the message: INFO IBM-MAIN

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN


Re: How should I send file to another sysplex securely.

2021-07-22 Thread Charles Mills
@Gil, works better than that. Watch my presentation  referenced earlier.

Charles


-Original Message-
From: IBM Mainframe Discussion List [mailto:IBM-MAIN@LISTSERV.UA.EDU] On Behalf 
Of Paul Gilmartin
Sent: Thursday, July 22, 2021 2:18 PM
To: IBM-MAIN@LISTSERV.UA.EDU
Subject: Re: How should I send file to another sysplex securely.

On Thu, 22 Jul 2021 14:05:31 -0600, Grant Taylor wrote:

>On 7/22/21 12:49 PM, Mike Hochee wrote:
>>...
>There is also a hybrid approach in which a symmetric key is used to
>encrypt / decrypt the data and asymmetric keys to protect the first key.
>  --  My understanding is that symmetric encryption is multiple orders
>of magnitude faster than asymmetric encryption.
> 
This is routinely, almost  universally, done for asymmetric encryption.
It lacks authentication and does not prevent MITM attacks:

o An intruder can masquerade as the sender and supply forged data.

o An intruder can masquerade as the recipient and intercept sensitive data.

o Or both, if you're lucky.

I believe (I'm mostly guessing) that a Certificate Authority provides
authentication in a repository of public keys but, "Quis custodiet ipsos
custodes?"  Computers come with a table of recognized CAs and their
public keys embedded in the OS.  This amounts to making the computer
vendors the ultimate Certificate authorities.

https://www.cs.cmu.edu/~rdriley/487/papers/Thompson_1984_ReflectionsonTrustingTrust.pdf

Or, the CAs could announce their public keys on NewsMax or Twitter.

>1)  Create an asymmetric public + private key pair on the destination
>system.
>2)  Transfer the destination system's public key to the source system.
>3)  Create a symmetric key on the source system.
>4)  Use the source system's symmetric key to encrypt the data.
>5)  Use the destination system's asymmetric public key to encrypt the
>source system's symmetric key.
>6)  Transfer both the encrypted data and the encrypted symmetric key
>from the source system to the destination system.
>7)  Use the destination system's asymmetric private key to decrypt the
>source system's symmetric key.
>8)  Use the decrypted source system's symmetric key to decrypt the data.
>...
>n)  PROFIT!!!
>
>The data and the symmetric key protecting it are only unencrypted on the
>source and destination system.

-- gil

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN


Re: How should I send file to another sysplex securely.

2021-07-22 Thread Paul Gilmartin
On Thu, 22 Jul 2021 14:05:31 -0600, Grant Taylor wrote:

>On 7/22/21 12:49 PM, Mike Hochee wrote:
>>...
>There is also a hybrid approach in which a symmetric key is used to
>encrypt / decrypt the data and asymmetric keys to protect the first key.
>  --  My understanding is that symmetric encryption is multiple orders
>of magnitude faster than asymmetric encryption.
> 
This is routinely, almost  universally, done for asymmetric encryption.
It lacks authentication and does not prevent MITM attacks:

o An intruder can masquerade as the sender and supply forged data.

o An intruder can masquerade as the recipient and intercept sensitive data.

o Or both, if you're lucky.

I believe (I'm mostly guessing) that a Certificate Authority provides
authentication in a repository of public keys but, "Quis custodiet ipsos
custodes?"  Computers come with a table of recognized CAs and their
public keys embedded in the OS.  This amounts to making the computer
vendors the ultimate Certificate authorities.

https://www.cs.cmu.edu/~rdriley/487/papers/Thompson_1984_ReflectionsonTrustingTrust.pdf

Or, the CAs could announce their public keys on NewsMax or Twitter.

>1)  Create an asymmetric public + private key pair on the destination
>system.
>2)  Transfer the destination system's public key to the source system.
>3)  Create a symmetric key on the source system.
>4)  Use the source system's symmetric key to encrypt the data.
>5)  Use the destination system's asymmetric public key to encrypt the
>source system's symmetric key.
>6)  Transfer both the encrypted data and the encrypted symmetric key
>from the source system to the destination system.
>7)  Use the destination system's asymmetric private key to decrypt the
>source system's symmetric key.
>8)  Use the decrypted source system's symmetric key to decrypt the data.
>...
>n)  PROFIT!!!
>
>The data and the symmetric key protecting it are only unencrypted on the
>source and destination system.

-- gil

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN


Re: How should I send file to another sysplex securely.

2021-07-22 Thread Charles Mills
@Mike, thanks for the kind words.

@Grant, yes, exactly, "pure" public key is way too slow to use for a 
significant file. The only practical approach is what you call "hybrid."

I would say in no event does the OP want to "roll his own" or "cobble something 
together out of bits and pieces."

This problem is what FTP does for a living. Either SFTP ("SSH FTP") which I am 
not real familiar with but I know works like a champ, or FTPS (FTP over TLS) 
which I am very familiar with. The two are totally different options; they do 
not interoperate. A client for one does not talk to a server for the other.

You need a secure (SFTP or FTPS) server at one end (sending or receiving, your 
choice) and some configuration at the other end. An investment in secure FTP is 
an investment in the future, not just this one problem.

You should be able to find lots of SHARE sessions and so forth on "how to 
install and configure and use a secure FTP server." There is help on this forum 
if necessary.

Oh! In Step 3 below, add to the sentence "... using a secure 
cryptographic-quality random-number generator." Again, you don't want to roll 
your own on this. Wy too many traps for the unwary.

Charles


-Original Message-
From: IBM Mainframe Discussion List [mailto:IBM-MAIN@LISTSERV.UA.EDU] On Behalf 
Of Grant Taylor
Sent: Thursday, July 22, 2021 1:06 PM
To: IBM-MAIN@LISTSERV.UA.EDU
Subject: Re: How should I send file to another sysplex securely.

On 7/22/21 12:49 PM, Mike Hochee wrote:
> With private key (aka symmetric ) the same key is used to encrypt and 
> decrypt, and the key must be securely shared among business partners 
> (a vulnerability).  Pervasive or z/OS data set encryption uses private 
> key encryption.
> 
> With public key model (aka asymmetric) a key pair is generated 
> and the keys are mathematically related, this enables the secure 
> sharing of a public key with another organization. Public key 
> cryptography is quite elegant IMO and solves your chicken/egg 
> issue.

There is also a hybrid approach in which a symmetric key is used to 
encrypt / decrypt the data and asymmetric keys to protect the first key. 
  --  My understanding is that symmetric encryption is multiple orders 
of magnitude faster than asymmetric encryption.

1)  Create an asymmetric public + private key pair on the destination 
system.
2)  Transfer the destination system's public key to the source system.
3)  Create a symmetric key on the source system.
4)  Use the source system's symmetric key to encrypt the data.
5)  Use the destination system's asymmetric public key to encrypt the 
source system's symmetric key.
6)  Transfer both the encrypted data and the encrypted symmetric key 
from the source system to the destination system.
7)  Use the destination system's asymmetric private key to decrypt the 
source system's symmetric key.
8)  Use the decrypted source system's symmetric key to decrypt the data.
...
n)  PROFIT!!!

The data and the symmetric key protecting it are only unencrypted on the 
source and destination system.



-- 
Grant. . . .
unix || die

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN


Re: How should I send file to another sysplex securely.

2021-07-22 Thread Grant Taylor

On 7/22/21 12:49 PM, Mike Hochee wrote:
With private key (aka symmetric ) the same key is used to encrypt and 
decrypt, and the key must be securely shared among business partners 
(a vulnerability).  Pervasive or z/OS data set encryption uses private 
key encryption.


With public key model (aka asymmetric) a key pair is generated 
and the keys are mathematically related, this enables the secure 
sharing of a public key with another organization. Public key 
cryptography is quite elegant IMO and solves your chicken/egg 
issue.


There is also a hybrid approach in which a symmetric key is used to 
encrypt / decrypt the data and asymmetric keys to protect the first key. 
 --  My understanding is that symmetric encryption is multiple orders 
of magnitude faster than asymmetric encryption.


1)  Create an asymmetric public + private key pair on the destination 
system.

2)  Transfer the destination system's public key to the source system.
3)  Create a symmetric key on the source system.
4)  Use the source system's symmetric key to encrypt the data.
5)  Use the destination system's asymmetric public key to encrypt the 
source system's symmetric key.
6)  Transfer both the encrypted data and the encrypted symmetric key 
from the source system to the destination system.
7)  Use the destination system's asymmetric private key to decrypt the 
source system's symmetric key.

8)  Use the decrypted source system's symmetric key to decrypt the data.
...
n)  PROFIT!!!

The data and the symmetric key protecting it are only unencrypted on the 
source and destination system.




--
Grant. . . .
unix || die

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN


Re: How should I send file to another sysplex securely.

2021-07-22 Thread Mike Hochee
Hi Colin, 

Best way probably varies depending on the use case, but basically you have two 
choices; public or private key cryptography. With private key (aka symmetric ) 
the same key is used to encrypt and decrypt, and the key must be securely 
shared among business partners (a vulnerability).  Pervasive or z/OS data set 
encryption uses private key encryption. 

With public key model (aka asymmetric) a key pair is generated and the keys are 
mathematically related, this enables the secure sharing of a public key with 
another organization. Public key cryptography is quite elegant IMO and solves 
your chicken/egg issue. There are many more  moving parts (see the latest draft 
of RFC 4880 for a look under the OPGP hood) , but most implementations do a 
good job of hiding  extraneous stuff. OpenPGP and OpenSSL are crypto systems 
based on the public key model. Secure email systems are typically use one of 
these.  A decent public key intro lives here 
https://en.wikipedia.org/wiki/Public-key_cryptography . Charles Mills gave an 
xlnt presentation which included coverage of the public key model in Oct 
2020...  https://www.newera-info.com/CM1.html 

HTH, 
Mike  

-Original Message-
From: IBM Mainframe Discussion List [mailto:IBM-MAIN@LISTSERV.UA.EDU] On Behalf 
Of Colin Paice
Sent: Thursday, July 22, 2021 10:08 AM
To: IBM-MAIN@LISTSERV.UA.EDU
Subject: How should I send file to another sysplex securely.

Caution! This message was sent from outside your organization.

I was wondering the best way customers send sensitive data between z/OS images.
I was thinking about exporting one's private certificates.

   1. I can create a dataset of the private certificates on system 1 and
   have it encrypted.   I can send it to the other system.   How can I decrypt
   it on the remote system as it needs shared certificates?  It seems a
   chicken and egg problem
   2. I can put a password on the file through JCL and use FTPS to send
   it.   This could easily be broken

This is hypothetical, but I would be interested in how to do it.

Colin Paice

--
For IBM-MAIN subscribe / signoff / archive access instructions, send email to 
lists...@listserv.ua.edu with the message: INFO IBM-MAIN

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN


Re: [EXTERNAL] Re: How should I send file to another sysplex securely.

2021-07-22 Thread Paul Gilmartin
On Thu, 22 Jul 2021 14:19:41 +, Horne, Jim wrote:

>Why wouldn't you just write a batch job to invoke SFTP?  It is z/OS to z/OS 
>and can handle almost all files, as far as I know
> 
Like most of the suggestions so far, this begs the question of transferring the 
key.

For "almost all files" you need either Dovetailed of flattening with TSO 
TRANSMIT.

Does sftp support anything similar to Diffie-Hellman key exchange?  That would 
still
leave the problem of authentication.

o Sneakernet?

o Cloud?  Still the authentication question.

-- gil

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN


Re: [EXTERNAL] Re: How should I send file to another sysplex securely.

2021-07-22 Thread Horne, Jim
Why wouldn't you just write a batch job to invoke SFTP?  It is z/OS to z/OS and 
can handle almost all files, as far as I know

Jim Horne

-Original Message-
How about using sftp - of course you would need to copy the file to an omvs 
file to do it, or get the Dovetail enhanced sftp which supports z/OS datasets.

NOTICE: All information in and attached to the e-mails below may be 
proprietary, confidential, privileged and otherwise protected from improper or 
erroneous disclosure. If you are not the sender's intended recipient, you are 
not authorized to intercept, read, print, retain, copy, forward, or disseminate 
this message. If you have erroneously received this communication, please 
notify the sender immediately by phone (704-758-1000) or by e-mail and destroy 
all copies of this message electronic, paper, or otherwise. By transmitting 
documents via this email: Users, Customers, Suppliers and Vendors collectively 
acknowledge and agree the transmittal of information via email is voluntary, is 
offered as a convenience, and is not a secured method of communication; Not to 
transmit any payment information E.G. credit card, debit card, checking 
account, wire transfer information, passwords, or sensitive and personal 
information E.G. Driver's license, DOB, social security, or any other 
information the user wishes to remain confidential; To transmit only 
non-confidential information such as plans, pictures and drawings and to assume 
all risk and liability for and indemnify Lowe's from any claims, losses or 
damages that may arise from the transmittal of documents or including 
non-confidential information in the body of an email transmittal. Thank you.

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN


Re: How should I send file to another sysplex securely.

2021-07-22 Thread ITschak Mugzach
encrypt and send over NJE using xmit?

ITschak Mugzach
*|** IronSphere Platform* *|* *Information Security Continuous Monitoring
for z/OS, x/Linux & IBM I **| z/VM coming soon  *




On Thu, Jul 22, 2021 at 5:08 PM Colin Paice  wrote:

> I was wondering the best way customers send sensitive data between z/OS
> images.
> I was thinking about exporting one's private certificates.
>
>1. I can create a dataset of the private certificates on system 1 and
>have it encrypted.   I can send it to the other system.   How can I
> decrypt
>it on the remote system as it needs shared certificates?  It seems a
>chicken and egg problem
>2. I can put a password on the file through JCL and use FTPS to send
>it.   This could easily be broken
>
> This is hypothetical, but I would be interested in how to do it.
>
> Colin Paice
>
> --
> For IBM-MAIN subscribe / signoff / archive access instructions,
> send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN
>

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN


Re: How should I send file to another sysplex securely.

2021-07-22 Thread Lionel B. Dyck
How about using sftp - of course you would need to copy the file to an omvs 
file to do it, or get the Dovetail enhanced sftp which supports z/OS datasets.


Lionel B. Dyck <><
Website: https://www.lbdsoftware.com
Github: https://github.com/lbdyck

“Worry more about your character than your reputation. Character is what you 
are, reputation merely what others think you are.”   - - - John Wooden

-Original Message-
From: IBM Mainframe Discussion List  On Behalf Of 
Colin Paice
Sent: Thursday, July 22, 2021 9:08 AM
To: IBM-MAIN@LISTSERV.UA.EDU
Subject: How should I send file to another sysplex securely.

I was wondering the best way customers send sensitive data between z/OS images.
I was thinking about exporting one's private certificates.

   1. I can create a dataset of the private certificates on system 1 and
   have it encrypted.   I can send it to the other system.   How can I decrypt
   it on the remote system as it needs shared certificates?  It seems a
   chicken and egg problem
   2. I can put a password on the file through JCL and use FTPS to send
   it.   This could easily be broken

This is hypothetical, but I would be interested in how to do it.

Colin Paice

--
For IBM-MAIN subscribe / signoff / archive access instructions, send email to 
lists...@listserv.ua.edu with the message: INFO IBM-MAIN

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN


How should I send file to another sysplex securely.

2021-07-22 Thread Colin Paice
I was wondering the best way customers send sensitive data between z/OS
images.
I was thinking about exporting one's private certificates.

   1. I can create a dataset of the private certificates on system 1 and
   have it encrypted.   I can send it to the other system.   How can I decrypt
   it on the remote system as it needs shared certificates?  It seems a
   chicken and egg problem
   2. I can put a password on the file through JCL and use FTPS to send
   it.   This could easily be broken

This is hypothetical, but I would be interested in how to do it.

Colin Paice

--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN