08-10-2016 06:06 AM
I'm sorry but I don't have access to Customer case and generally I dont't have access to brocade Intranet.
My question "if you have opened a TAC" was related just to know what brocade say accord you post from:
07-04-2016 02:52 PM
>It looks like a software issue on the ADX. Brocade are investigating...
--->>> they say they opened the TAC with Brocade on Jun 28, 2016.
this is very courious, Brocade Support is usually very quickly with answer.
I would suggest, you or the Ticket owner to send a reminder by firstname.lastname@example.org
08-12-2016 09:32 AM
It might be worth asking Rackspace where Brocade is in the diagnosis of the problem.
In general I see a few steps, eg acknowledgement/understanding of the issue, replication of the issue, escalation into L3, characterising/opening a DEFECT and then engineering working on a fix with subsequent testing.
As part of this they may also come up with a workaround.
If you don't know which part of the process the case is in then I can see it can be very frustrating, as it looks like no progress has been made. Also with intermittent problems it can take a while to adequately simulate and reproduce the problem at will. For example if only a small percentage of transactions cause a problem then Brocade would need to figure out whether there is something within these transactions that causes a problem or whether there is something else going on in the ADX that causes some transactions to not work.
Having a third party between you and TAC can also be challenging as you are relying on them passing information accurately and in a timely manner between you and TAC.
08-15-2016 03:04 AM
Just got an update from Brocade (via RackSpace) this morning:
Our engineering team is currently investigating this issue. When the issue occurs, the CSW debug shows the data marked as complete after the header and performs the forward action hence the post data is lost. This issue is being tracked under new defect DEFECT000610535.
The temporary work around is to disable the CSW policy for the port as then the issue is not seen. This will disabled the client IP insertion the policy is used for.
So confirmed as a software defect. Will let you know when we get a timeline for a fix.