Re: QAUDJRN

Le 04/05/2022 à 16:34, Greg Wilburn a écrit :

Well Jim, 99.99% of the authority failures are IBM, period. And they're not doing a very good job of figuring it out.

The offender? QWQADMIN (web query admin)
The object(s)? '/QIBM/UserData/qwebqry/apps/base_synonyms'

Even if it is probably part of WebQuery product, I have seen many times that authority was broken by restore operations. I have seen it many times on /QIBM/ProdData/HTTPA directory, which retains Navigator for i web application files, for instance. Every time, analysing the Authority Failure audit entries was a huge help to solve the issue.

What is the detail of the AF entries? Are they all related to the same object, to the same user profile? What is the violation type?

This can help to properly setup the required authority.

Another way to solve such issues that I used in the past was to CHKPRDOPT or SAVLICPGM/RSTLICPGM the product. Depending on the product, there are sometimes exit programs invoked by those commands which reset properly the authority.

It will be my hide when the system goes down because all of the disk space is consumed.

So until IBM support can "fix" this (which they haven't been able up to this point), it's going to have to stay off.
We don't have an audit or auditor.

I will still pursue... but if IBM can't figure it out, what am I supposed to do?

-----Original Message-----
From: MIDRANGE-L <midrange-l-bounces@xxxxxxxxxxxxxxxxxx> On Behalf Of Jim Oberholtzer
Sent: Tuesday, May 3, 2022 5:34 PM
To: Midrange Systems Technical Discussion <midrange-l@xxxxxxxxxxxxxxxxxx>
Subject: Re: QAUDJRN

If you have authority failures, that’s a huge red flag. You need to find out why they are happening and correct it. No auditor worth the money is going to allow that.

It might be frustrating, but the tooling you have available should lead you to the answer. Look at the authority collection process to figure out who, what, and why that’s happening.

Ignore them at the peril of the system, the company, and your employment. Who’s going to get blamed when bad things happen, it won’t be the boss. Accepting the explanation as given is not acceptable….

Jim Oberholtzer
Agile Technology Architects

On May 3, 2022, at 3:20 PM, Marc Rauzier <marc.rauzier@xxxxxxxxx> wrote:

Le 03/05/2022 à 21:17, Greg Wilburn a écrit :

So I'm ready to turn this thing off... Actually, I already have.

I changed QAUDCTL to *NONE, but because my QAUDJRN was setup with MNGRCV(*SYSTEM) it creates a new receiver each morning when we IPL.

That's the normal behavior of journal management. You should only see entries related to journal and receivers management, such as new receiver created and attached, previous receiver detached, journal and receivers backed up... This cannot use a lot of disk space. If you specify DLTRCV(*YES) for AUDJRN journal, you will have only one receiver in the chain, even if it is not really a recommended setup for QAUDJRN journal. If you do no longer want to use audit (well, this is not really recommended nowadays) and do not want to manage the journal, you can delete it then delete the receivers.

BUT, according to IBM there shouldn’t be anything in the receiver file. There are definitely ZC entries.

Unless I am wrong, ZC/ZR entries come with QAUDCTL system value set with *OBJAUD. So that is quite strange if they are still here when QAUDCTL is set to *NONE. Are you sure that those ZC entries are *after* QAUDCTL was set to *NONE?

I'm so frustrated right now... why would IBM give you something that has the potential to EAT DISK SPACE at an alarming rate, yet not provide any "cleanup" tools that aren't supplied "AS-IS". Yet we are supposed to turn this on??

Audit journal needs to be properly managed. Normally, you may want to backup then delete the receivers and keep the backup long enough to comply with security rules which require the ability to analyse them in case of security failure. If you do not want to create tools for that, I suggest you to look at some commercial software. Security becomes more and more a key topic nowadays.

FWIW, the DB2 Web Query folks came back and said to turn off auditing on the directory (and subtree) that is generating 99.7% of the AF entries.

As far as I know, you cannot do that. There is no directory or file or object attribute which prevents Authority Failures entries in audit journal. As soon as QAUDLVL(2) system value includes *AUTFAIL, the system will log *all* authority failures. This will always be a matter for security administrator to solve them with setting up appropriate authority. My 2 cents.

Is this really worth all the effort?

--
This is the Midrange Systems Technical Discussion (MIDRANGE-L) mailing list
To post a message email: MIDRANGE-L@xxxxxxxxxxxxxxxxxx
To subscribe, unsubscribe, or change list options,
visit: https://lists.midrange.com/mailman/listinfo/midrange-l
or email: MIDRANGE-L-request@xxxxxxxxxxxxxxxxxx
Before posting, please take a moment to review the archives
at https://archive.midrange.com/midrange-l.

Please contact support@xxxxxxxxxxxxxxxxxxxx for any subscription related questions.

Help support midrange.com by shopping at amazon.com with our affiliate link: https://amazon.midrange.com