- Local User Groups
There are two different installation ISOs for R80.20:
So yes, you can upgrade a standalone to R80.20.
If you do an in place upgrade of an R80.10 management server to R80.20 will the kernel be updated or does this require a fresh build and migrate export/import?
Kernel is upgraded on in-place upgrades, but filesystem change from ext3 to xfs is not carried out. This requires advanced upgrade (export/import).
Hi, I was going to upgrade from r80.10 to r80.20 soon. You mentioned the "ext3 to xfs is not carried out" is this required and how to access advanced upgrade (export/import)?
In order to exploit this, you'd have to be an authorized user (most likely an admin of some sort) and get to expert mode (i.e. not be in clish).
That assumes the bug is present in our kernel, which I can't say for sure.
In any case, these sorts of issues tend to get addressed in later releases/jumbo hotfixes since they pose no immediate risk.
Of course, this is just my initial assessment and should not be construed as official.
My guess is an official SK will appear on this in the near future.
If it's urgent, I recommend opening a TAC case.
Clarification: future of HCL (hardware compatibility):
Open server compatibility is expected to be much better with the new linux release.
As i explained above, mgmt GA works w the updated linux (so it is integrated in the GA). Updated HCL was published (mgmt is user space and therefore simpler).
We also have EA with r80.20 variant based on updated linux for GA - you are welcome to join this EA asap. With this release we will have the latest open servers supported in timely manner (so we hope to put to delayed HCL behind us, once we release the variant that works on the new linux).
Thanks for the reply Dorit, is this GW EA also for VSX?
Within the open servers for HCL are you testing the new 25G nic together with the new servers?
1. The gw variant for new linux is in production EA for both open servers and clous usage
2. VSX w new linux is developed but not yet ready for EA. Why? With the gw move to new linux, the biggest change is replacing the VFRF module with “Name spaces” as VFRF no longer exist. So we released Gw-non-VSX first and now we are QAing the VSX. The exact date of EA for VSX depends on our quality validation and is expected to happen this quarter.
If you will join the EA for the new kernel GW, you will also get updates on the missing functionality (vsx)
Last: The exact functionality we certified for open server is based on priorities from the field (at the end we want to cerify everything possible and we believe that w the new kernel the time to market and quality of open servers support will be dramatically better and we can support anything needed).
To impact the order/priority of certifications for HCL, contact the local field and get them to present the business case to our solution center (the entity that manages the field priritization to such things)
No, there is a GA for gateway with R80.20 with kernel 2.6 (same kernel as R80.10) which is fully supported.
In parallel there is an EA for gateway with R80.20 with kernel 3.10 (same kernel as R80.20 Mgmt GA and R80.20.M1), which has limited scope.
Cluster object name still cannot be the same as policy package name (and vice versa). Is there any special reason for that ? Beside that the name will be the same ?
yep, this also happens in case you are going to create for example new Network Group with the name exactly the same as something else ...
And to use a clear naming convention is useful for ages now (and so you don't run is such problems)
What's New in v1.3
Hello, do you have any news about this new feature indicated in r80.20 EA?
There isn't in r80.20 GA?
Hello, thanks for reply.
This is the bug fix note:
The "Categorize HTTPS sites" feature does not work when HTTPS Inspection is enabled. Refer to sk90840.
The sk indicated (sk90840) isn't linked to this issue:
sk90840 - HTTPS Inspection is not supported for IPv6 traffic in R76 / R77.X / R80.10
Can you verify?