AsyncO/S 7.6.0-444

[David Owens]

I am currently piloting the FCS 7.6.0-444 release and have found that it does appear to fix issues we are seeing with 7.5.1-028 on our production ESA's.  I am inquiring about the stability of the 7.6.0-444 release, how many have installed it and what new issues have been reported with the FCS?  Also any update on when this release will be going to GA?

[Donald Nash]

We're at 7.6.0-444. We encountered a bug that causes messages to vanish if three conditions are met: 1) the message is received on a private listener, 2) the message is splintered due to an outgoing policy, and 3) the message is to have a DKIM signature applied. I was told that it's due to be fixed in 7.6.1. Fortunately, we were able to work around it by reducing our use of outgoing policies to avoid message splintering, but that was only the case because our outgoing policies were pretty simple to begin with.

++Don

[David Owens]

Don,

Have you ever heard anything else on your splintering issue above?  This does not affect me in the pilot environment as we do not enforce DKIM in that environment, but will have a major effect when I move to production.  I am getting concerned that 7.6.x  id going to be buggy.

David

[Donald Nash]

David,

I haven't pursued the matter any further. As I said, I was told that the fix will be in 7.6.1, which doesn't appear to have been released yet. The ETA we were given has not yet arrived, but it's close. I'm waiting patiently... Fortunately, the work-around we used has been completely effective and hasn't caused any undesirable side effects.

As for 7.6 being buggy, keep in mind that it hasn't reached GA yet. GA is still 7.5.1-102. I'd wait for 7.6 to be promoted to GA before making any judgements about its bugginess.

++Don

[Platform Team]

I also encountered this bug in 7.6.0-444 when we upgraded recently. I was told a workaround is to use public instead of private listeners, but this didn't fix the issue for us (and took some time to set up - which was wasted).

I too was told the bug is fixed in 7.6.1, but have not been able to get a release date for that version. In the end, we had to rollback to 7.5.2.

Interesting to note that Cisco support acknowedge this defect but it is not in the release notes as a known issue.

[Donald Nash]

Ross,

We were also given the "switch to a public listener" advice, which I decided was too much work and too much risk for a bug that we could deal with in easier ways. It's very interesting that the bug has continued to bite you, since the bug description specifically says that it only happens on private listeners. I suggest opening a support case and pushing them to dig into it. Perhaps they don't understand the bug as well as they think they do, or perhaps this is a different bug.

The fact that it isn't in the release notes probably means it wasn't discovered until after they went to press.

++Don

[robertmibus]

Hi Don,

We tested 7.6.0-444 extensively before moving it in to production, and all seemed fantastic. We moved it into production, and are making good use of some of the new features.

After about a week in production, we've encountered an issue where certain mis-formed recipients can cause the work queue to get stuck; we've been told it'll be resolved in 7.6.1, and we have a semi-workaround in place - but it's not ideal if you have potentially untrusted hosts with relay access.

(We're an ISP, so we have several hundred thousand customer IPs with relay access).

It's not enough for us to want to roll production back to 7.5, but I'm certainly very much looking forward to 7.6.1.

I'd probably consider waiting for 7.6.1 if you can, but it really depends on the level of discomfort you're have with 7.5.

HTH;

Robert.

[Donald Nash · Accepted solution]

Robert,

I think you were addressing David, not me.

I agree with your sentiment. I recommend sitting out 7.6.0 and waiting for whatever 7.6.x is declared GA. That is, unless you really need some of the new features it has.

++Don

[David Owens]

Robert,

I apologize for the "buggy' comment as I do realize that the current iteration of 7.6.0 is beta code.  It was not my intent to cast doubt on the upcoming GA release of 7.6.1 code as my experience in the past is that GA code has been reliable when delivered.

Don and Robert,

I appreciate both of your input regarding waiting for the 7.6.1 release, it is our current plan to now wait.  We are frustrated with the Office file scan filter issue in version 7.5.1 and need a solution as soon as possible to meet our business requirements. 

The input on the message splintering and mis-formed recipients has assisted our decision process to wait for the 7.6.1 release, I have really appreciated the input from both of you regarding your experince wwith 7.6.0.  We have installed 7.6.0 on our pilot environment ESA's but due to environment restrictions we have not identified the issues you reported above (by design our pilot environment does not fully mirror our production environment).

Thanks everyone for your input,

David

[Donald Nash]

I wasn't offended, so no apology is necessary. FYI, 7.6.0 isn't beta, it's FCS, the middle step between beta and GA.

++Don

[David Owens]

Sorry – not use to all the terms. Thanks for understanding.

Thanks,

David Owens