[Beowulf] RHEL7 kernel update for L1TF vulnerability breaks RDMA

[Jonathan Engwall]

I am not shocked that my previous message may have been removed. There is a bit of brute science involved as well. A false burst may seem to tune a network. A properly timed burst is more effective in gaining authorization. You don't need magic code to leverage a network you just need to generate request that is swifter to answer than drop.
Your own configurations can be to blame. Suppose you require a percentage of answered to blocked requests. You open that door.
Suppose you want a specific set up time, whatever it is the shorter the better. When those connections go underutilized you may have a smoking gun.

Any fans of NPTEL? A lecture I watched today was quite synchronised.

Earlier I should have written so NOT do this...

You would also be looking for many many symlinks  for users who exploit the network itself, I am on!y trying to help and don't know for a fact or do any of these sorts of things.

On August 18, 2018, at 4:57 AM, Chris Samuel <chris at csamuel.org> wrote:

On 18/8/18 8:47 pm, Jörg Saßmannshausen wrote:

> Hi Chris,

Hiya,

> these are bad news if InfiniBand will be affected here as well as
> that is what we need to use for parallel calculations. They make use
> of RMDA and if that has a problem..... well, you get the idea I
> guess.

Oh yes, this is why I wanted to bring it to everyones attention, this
isn't just about Lustre, it's much more widespread.

> Has anybody contacted the vendors like Mellanox or Intel regarding
> this?

As Kilian wrote in the Lustre bug quoting his RHEL bug:

     https://bugzilla.redhat.com/show_bug.cgi?id=1618452

     — Comment #3 from Don Dutile <ddutile at redhat.com> —
     Already reported and being actively fixed.

     Cannot make this public, as the patch that caused it was due to 
embargo'd
     security fix.

     This issue has highest priority for resolution.
     Revert to 3.10.0-862.11.5.el7 in the mean time.

     This bug has been marked as a duplicate of bug 1616346

-- 
  Chris Samuel  :  http://www.csamuel.org/  :  Melbourne, VIC
_______________________________________________
Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin Computing
To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf