Hi Mahesh,
Ack with comments, [Vu].
Regards, Vu
> -Original Message-
> From: mahesh.va...@oracle.com [mailto:mahesh.va...@oracle.com]
> Sent: Friday, February 24, 2017 10:34 AM
> To: canh.v.tru...@dektech.com.au; vu.m.ngu...@dektech.com.au;
> lennart.l...@ericsson.com
> Cc:
Hi Hans,
With one hash and one crypto function, I would rather go with our own
implementation. Both of them is not hard to implement.
If there is a need for supporting more algorithms in the future, then I would
consider for using OpenSSL.
As Anders has pointed out, adding OpenSSL means adding
Hi Vu,
ACK with following for All patch list
Note : default functionality tested , Enhancement specific
functionality not tested.
- consider optimizing encoding function uint32_t edp_ed_open_stream_rec()
- Update PR document
- Is multiple alternative destinations allowed concurrently ?
Good catch. Hitting the case of fm_peer_down_wait() is very unlikely.
But here fm_peer_down_wait() is called only before fm nid_notifies and
considering for amfnd-up event too. A rare and race condition can hit
where fm on upcoming new active receives fm-down event and amfnd is
still alive.
Hi Ramesh,
One minor query:
In RED_UP of peer AVD, newly active SC will reboot itself if peer FM on
old active SC is not up. If this true then in which situations newly
active SC will wait in fm_peer_down_wait().
Thanks,
Praveen
On 22-Feb-17 5:00 PM, ramesh.bet...@oracle.com wrote:
>
Hi Vu,
Except replacing `const auto`, incorporated all other comments.
I will consider replacing `const auto` after 5.2 tag, to avoid last
minute change .
-AVM
On 2/24/2017 1:52 PM, Vu Minh Nguyen wrote:
> Hi Mahesh,
>
> Ack with comments, [Vu].
>
> Regards, Vu
>
>> -Original
On 24-Feb-17 4:07 PM, ramesh betham wrote:
> Good catch. Hitting the case of fm_peer_down_wait() is very unlikely.
>
> But here fm_peer_down_wait() is called only before fm nid_notifies and
> considering for amfnd-up event too. A rare and race condition can hit
> where fm on upcoming new active
HI Mahesh,
> - Is multiple alternative destinations allowed concurrently ?
No. With current design, streaming log records to multiple Unix Domain
Socket destinations is done consecutively.
Regards, Vu
> -Original Message-
> From: A V Mahesh [mailto:mahesh.va...@oracle.com]
> Sent:
Hi Mahesh
Ack with comments
Comments inline [Lennart]
* Remove '#include "base/ncspatricia.h"' in lgs_stream.h. Not needed
* You could use C++ 11 range based for loops to simplify iterations. See
comment/example below. [Lennart]
* This patch do pass legacy tests but the following cpp check
I don't think RED_UP always come first or later MDS_UP or vice-versa.
Most likley it depends on the sequence of svc registrations happens with
MDS.
Thanks,
Ramesh.
On 2/24/2017 4:48 PM, praveen malviya wrote:
>
>
> On 24-Feb-17 4:07 PM, ramesh betham wrote:
>> Good catch. Hitting the case of
Hi Zoran,
yes, I also think Anders solution is ok. Later we may consider including
OpenSSL if needed.
/Regards HansN
On 02/24/2017 10:23 AM, Zoran Milinkovic wrote:
> Hi Hans,
>
> With one hash and one crypto function, I would rather go with our own
> implementation. Both of them is not hard
Summary: amf: add support for restrictions to auto-repair
Review request for Trac Ticket(s): 2144
Peer Reviewer(s): praveen
Pull request to:
Affected branch(es): default
Development branch:
Impacted area Impact y/n
Docs
src/amf/amfd/ndproc.cc | 6 --
1 files changed, 4 insertions(+), 2 deletions(-)
This patch adds support for Section 3.11.1.4.2 of AMF B.04.01 spec: Restrictions
to Auto-Repair.
diff --git a/src/amf/amfd/ndproc.cc b/src/amf/amfd/ndproc.cc
--- a/src/amf/amfd/ndproc.cc
+++
13 matches
Mail list logo
