On Thu, Jun 14, 2018 at 10:19 PM Richard W.M. Jones <rjones@redhat.com> wrote:
On Thu, Jun 14, 2018 at 09:16:01PM +0300, Nir Soffer wrote:
> When sending request with small or no payload, it is simpler and
> possibly more efficient to use the high level HTTPSConnection.request(),
> instead of the lower level APIs.
>
> The only reason to use the lower level APIs is to avoid copying the
> payload, or on python 2, to use a bigger buffer size when streaming a
> file-like object.

Change seems quite straightforward.  I'll have to test this out on my
own oVirt system first before I can properly verify it.

I tested the same change with imageio example upload script, and 
it *doubled* the throughput.

See https://gerrit.ovirt.org/#/c/92275/

Thanks, Rich.

>  v2v/rhv-upload-plugin.py | 35 ++++++++++++++++-------------------
>  1 file changed, 16 insertions(+), 19 deletions(-)
>
> diff --git a/v2v/rhv-upload-plugin.py b/v2v/rhv-upload-plugin.py
> index ed99cc7a9..3fad865f6 100644
> --- a/v2v/rhv-upload-plugin.py
> +++ b/v2v/rhv-upload-plugin.py
> @@ -263,12 +263,12 @@ def pread(h, count, offset):
>      transfer = h['transfer']
>      transfer_service = h['transfer_service']

> -    http.putrequest("GET", h['path'])
> +    headers = {"Range", "bytes=%d-%d" % (offset, offset+count-1)}
>      # Authorization is only needed for old imageio.
>      if h['needs_auth']:
> -        http.putheader("Authorization", transfer.signed_ticket)
> -    http.putheader("Range", "bytes=%d-%d" % (offset, offset+count-1))
> -    http.endheaders()
> +        headers["Authorization"] = transfer.signed_ticket
> +
> +    http.request("GET", h['path'], headers=headers)

>      r = http.getresponse()
>      # 206 = HTTP Partial Content.
> @@ -319,11 +319,10 @@ def zero(h, count, offset, may_trim):
>                        'size': count,
>                        'flush': False}).encode()

> -    http.putrequest("PATCH", h['path'])
> -    http.putheader("Content-Type", "application/json")
> -    http.putheader("Content-Length", len(buf))
> -    http.endheaders()
> -    http.send(buf)
> +    headers = {"Content-Type": "application/json",
> +               "Content-Length", str(len(buf))}
> +
> +    http.request("PATCH", h['path'], body=buf, headers=headers)

>      r = http.getresponse()
>      if r.status != 200:
> @@ -368,11 +367,10 @@ def trim(h, count, offset):
>                        'size': count,
>                        'flush': False}).encode()

> -    http.putrequest("PATCH", h['path'])
> -    http.putheader("Content-Type", "application/json")
> -    http.putheader("Content-Length", len(buf))
> -    http.endheaders()
> -    http.send(buf)
> +    headers = {"Content-Type": "application/json",
> +               "Content-Length", str(len(buf))}
> +
> +    http.request("PATCH", h['path'], body=buf, headers=headers)

>      r = http.getresponse()
>      if r.status != 200:
> @@ -387,11 +385,10 @@ def flush(h):
>      # Construct the JSON request for flushing.
>      buf = json.dumps({'op': "flush"}).encode()

> -    http.putrequest("PATCH", h['path'])
> -    http.putheader("Content-Type", "application/json")
> -    http.putheader("Content-Length", len(buf))
> -    http.endheaders()
> -    http.send(buf)
> +    headers = {"Content-Type": "application/json",
> +               "Content-Length", str(len(buf))}
> +
> +    http.request("PATCH", h['path'], body=buf, headers=headers)

>      r = http.getresponse()
>      if r.status != 200:
> --
> 2.17.1

--
Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones
Read my programming and virtualization blog: http://rwmj.wordpress.com
virt-top is 'top' for virtual machines.  Tiny program with many
powerful monitoring features, net stats, disk stats, logging, etc.
http://people.redhat.com/~rjones/virt-top