[FFmpeg-devel] [PATCH] avformat: AMQP: add option delivery_mode

Andriy Gelman andriy.gelman at gmail.com
Sat Jun 20 08:36:14 EEST 2020


Hi Levis, 

Thanks for your patch.

On Mon, 15. Jun 11:16, Levis Florian wrote:
> Signed-off-by: Levis Florian <levis.florian at gmail.com>
> ---
>  doc/protocols.texi    | 13 +++++++++++++
>  libavformat/libamqp.c |  6 +++++-
>  2 files changed, 18 insertions(+), 1 deletion(-)
> 
> diff --git a/doc/protocols.texi b/doc/protocols.texi
> index 7aa758541c..336246e67a 100644
> --- a/doc/protocols.texi
> +++ b/doc/protocols.texi
> @@ -109,6 +109,19 @@ the received message may be truncated causing decoding errors.
>  The timeout in seconds during the initial connection to the broker. The
>  default value is rw_timeout, or 5 seconds if rw_timeout is not set.
>  
> + at item delivery_mode @var{mode}
> +Sets the delivery mode.
> +

> +The following values are recognized:
> + at table @samp
> + at item persistent
> +Delivery mode set to "persistent" (2). This is the default value.
> +
> + at item non-persistent
> +Delivery mode set to "non-persistent" (1)
> +
> + at end table
> +

Please add more details on trade-offs between modes.

>  @end table
>  
>  @section async
> diff --git a/libavformat/libamqp.c b/libavformat/libamqp.c
> index aaf0e51152..83de2229fb 100644
> --- a/libavformat/libamqp.c
> +++ b/libavformat/libamqp.c
> @@ -39,6 +39,7 @@ typedef struct AMQPContext {
>      int pkt_size;
>      int64_t connection_timeout;
>      int pkt_size_overflow;
> +    int delivery_mode;
>  } AMQPContext;
>  
>  #define STR_LEN           1024
> @@ -52,6 +53,9 @@ static const AVOption options[] = {
>      { "exchange", "Exchange to send/read packets", OFFSET(exchange), AV_OPT_TYPE_STRING, { .str = "amq.direct" }, 0, 0, .flags = D | E },
>      { "routing_key", "Key to filter streams", OFFSET(routing_key), AV_OPT_TYPE_STRING, { .str = "amqp" }, 0, 0, .flags = D | E },
>      { "connection_timeout", "Initial connection timeout", OFFSET(connection_timeout), AV_OPT_TYPE_DURATION, { .i64 = -1 }, -1, INT64_MAX, .flags = D | E},
> +    { "delivery_mode",  "Delivery mode", OFFSET(delivery_mode), AV_OPT_TYPE_INT, { .i64 = AMQP_DELIVERY_PERSISTENT }, 1, 2, .flags = E, "delivery_mode"},
> +    { "persistent",     "persistent delivery mode",     0, AV_OPT_TYPE_CONST, {.i64=AMQP_DELIVERY_PERSISTENT }, 0, 0, E, "delivery_mode" },
> +    { "non-persistent", "non-persistent delivery mode", 0, AV_OPT_TYPE_CONST, {.i64=AMQP_DELIVERY_NONPERSISTENT }, 0, 0, E, "delivery_mode" },
>      { NULL }

Is the goal of the non-persistent option to reduce disk io, or something else?
I ran a quick test to look at disk io, but didn't see a difference.

>  };
>  
> @@ -222,7 +226,7 @@ static int amqp_proto_write(URLContext *h, const unsigned char *buf, int size)
>  
>      props._flags = AMQP_BASIC_CONTENT_TYPE_FLAG | AMQP_BASIC_DELIVERY_MODE_FLAG;
>      props.content_type = amqp_cstring_bytes("octet/stream");
> -    props.delivery_mode = 2; /* persistent delivery mode */
> +    props.delivery_mode = s->delivery_mode;
>  
>      ret = amqp_basic_publish(s->conn, DEFAULT_CHANNEL, amqp_cstring_bytes(s->exchange),
>                               amqp_cstring_bytes(s->routing_key), 0, 0,
> -- 
> 2.27.0
> 
> _______________________________________________
> ffmpeg-devel mailing list
> ffmpeg-devel at ffmpeg.org
> https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
> 
> To unsubscribe, visit link above, or email
> ffmpeg-devel-request at ffmpeg.org with subject "unsubscribe".

Thanks,
-- 
Andriy


More information about the ffmpeg-devel mailing list