Yes, you are right. We need that heart-beat mechanism.
On Thu, Jan 12, 2017 at 11:43 AM, Angel Luis Jimenez Martinez <
soyan...@gmail.com> wrote:
> So if I understand correctly tcp user timeout applies to sent data that
> has not been acknowledged. So in this case maybe it would not apply
If the clients dies silently (without sending the TCP RST), the server side
connection will stay until TCP user time out is triggered (order of tens of
minutes depending on your kernel setup) in the current implementation
unless you have a custom dialer to turn on TCP KeepAlive. We are working
We will have heart-beat mechanism ready to tackle this kind of issues soon
(Q1 2017).
On Tue, Dec 13, 2016 at 10:32 AM, Paul Breslin
wrote:
> This seems to still be an issue. We have a server side stream read
> implementation that is forced to manually timeout because
ec 5, 2016 at 2:17 PM Qi Zhao <zh...@google.com> wrote:
>>>>
>>>> On Mon, Dec 5, 2016 at 12:13 PM, Arya Asemanfar <
>>>> arya.aseman...@mixpanel.com> wrote:
>>>>
>>>> Thanks for the feedback. Good idea re metadata for gettin
Arya Asemanfar <
>> arya.aseman...@mixpanel.com> wrote:
>>
>> Thanks for the feedback. Good idea re metadata for getting the Balancer
>> to treat the connections as different. Will take a look at that.
>>
>> Some clarifications/questions inline:
>>
>&g
...@mixpanel.com> wrote:
>>
>> Thanks for the feedback. Good idea re metadata for getting the Balancer
>> to treat the connections as different. Will take a look at that.
>>
>> Some clarifications/questions inline:
>>
>> On Mon, Dec 5, 2016 at 11:11 AM,
> On Mon, Dec 5, 2016 at 11:11 AM, 'Qi Zhao' via grpc.io <
> grpc-io@googlegroups.com> wrote:
>
>> Thanks for the info. My comments are inline.
>>
>> On Sun, Dec 4, 2016 at 7:27 PM, Arya Asemanfar <
>> arya.aseman...@mixpanel.com> wrote:
>>
>>>
On Tue, Oct 11, 2016 at 4:59 AM, wrote:
>
>
>
>>> moreover, what can grpc applications do to read from this channel ?
>>>
>> I actually already provided a solution to you in my first reply. Does it
>> work for you?
>>
>>>
>>> The solution you provided doesnt work for me. The
Your client code seems incomplete (there is no code sending rpcs). Can you
share the real code you used to get that latency numbers?
On Thursday, September 15, 2016 at 7:48:54 AM UTC-7,
ew...@grandslammedia.com wrote:
>
> I've attached a few pastebins of code samples:
>
> proto file: