ssl_client_session_cache size

[Charlie Morgan]

Hey Ingela,

Thanks for the reply, it's OTP 24.0.5.

Understood, I was hesitant to make an issue as I was struggling to recreate
although seeing it in deployment. I looked through the tests and I don't
think this section actually has test coverage, in the meantime we have
reduced the session cache lifetime to keep our memory down.

Thanks,

Charlie

On Mon, Aug 23, 2021 at 2:45 AM Ingela Andin <ingela.andin@REDACTED> wrote:

> Hi!
>
> Could you be a little more specific about the OTP-24 version. Is it 24.0
> or a patch version? Also it would be appropriate to create a github issue
> instead of reporting it here, where you  also provide a small recreation
> example.  As far as I know this has been an issue in the past but
> should work for OTP-24. Our tests are passing, so we really need something
> more to go on.
>
> Regards Ingela Erlang/OTP team - Ericsson AB
>
> Den lör 21 aug. 2021 kl 10:38 skrev Charlie Morgan <chargreezy@REDACTED>:
>
>> Hi!
>>
>> We have a lot of outgoing ssl connections and noticed that ssl_manager
>> does not respect the default session_cache_client_max value and the ets
>> session cache table increases unbounded until we hit the session lifetime.
>>
>> I have tried to write some tests to replicate this behavior but it's
>> proving difficult, do you know of any circumstances that could be causing
>> this? My initial thoughts were problems with
>> ssl_manager:do_register_session.
>>
>> Currently using OTP24
>>
>> Thanks!
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://erlang.org/pipermail/erlang-questions/attachments/20210823/cb34e8b6/attachment.htm>