Hi, 

Hard to say what happens there but it might be that ldp/vls (the shims on top 
of vcl) do not handle properly whatever nginx does when it tries to reload. In 
particular, if it involves an exec syscall, we probably lose state and can’t 
recover. 

As for the leak, what version of vpp/vcl are you using? Can you try with 
something recent like 22.02?

Regards,
Florin

> On Mar 10, 2022, at 7:41 PM, first_se...@163.com wrote:
> 
> hello ,sir, I have one question:
>       I used nginx as proxy with vcl, core dump was happened when i 
> reloadnginx with nginx -s reload。and i dgb the core file,the inforamtion  
> means Core was generated by `nginx: worker process is shutting'.
> Program terminated with signal SIGSEGV, Segmentation fault. and the stack 
> show vppcom_session_disconnect (session_handle=<optimized out>) at 
> /home/from_root/workspace/vpp/src/vcl/vppcom.c and 
> listen_session->n_accepted_sessions--;
> like this
> <1646969309.png>
> on the other hand,i found RAM of nginx workers always increase when wrk every 
> time and  not decrease when finished  wrk.
> at the time that beyound the totle memory。it also happens core dump and gdb 
> cannot show useful information.like this
> <企业微信截图_16469 701207851.png><企业微信截图_16469 69760375.png>
> 
> 
> 

-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.
View/Reply Online (#21004): https://lists.fd.io/g/vpp-dev/message/21004
Mute This Topic: https://lists.fd.io/mt/89703782/21656
Mute #vppcom:https://lists.fd.io/g/vpp-dev/mutehashtag/vppcom
Mute #vpp-dev:https://lists.fd.io/g/vpp-dev/mutehashtag/vpp-dev
Group Owner: vpp-dev+ow...@lists.fd.io
Unsubscribe: https://lists.fd.io/g/vpp-dev/unsub [arch...@mail-archive.com]
-=-=-=-=-=-=-=-=-=-=-=-

Reply via email to