wget retry problem/bug ver 1.10.2

Hi - I am having a problem with wget v1.10.2 when I run the command:

wget -t 100 -T 10 -i files.txt

On older versions (eg v1.6), this will keep retrying and will "resume"
the download from where it left off, EVERY time.

On the latest wget version, this will do the retry properly if the
connection times-out in the middle of receiving bytes (read-error),
however, if there is a timeout when receiving the header response from
the server, wget will start all over again and will not resume.

Here is a sample of the error:
-------------------
HTTP request sent, awaiting response... Read error (Connection timed
out) in headers.
------------------

I've looked at all the switches and can't figure out how to make this
work like the older versions. Does anybody have any solutions, or is
this a bug? BTW - I know there are other progs but I would first like
to see if there is a solution for wget.

Thanks.

(E-Mail Removed) wrote:
> Hi - I am having a problem with wget v1.10.2 when I run the command:
>
> wget -t 100 -T 10 -i files.txt
>
> On older versions (eg v1.6), this will keep retrying and will "resume"
> the download from where it left off, EVERY time.

From man page:
--tries=number
Set number of retries to number. Specify 0 or inf for infinite
retrying. The default is to retry 20 times, with the
exception of
fatal errors like ``connection refused'' or ``not found'' (404),
which are not retried.

> On the latest wget version, this will do the retry properly if the
> connection times-out in the middle of receiving bytes (read-error),
> however, if there is a timeout when receiving the header response from
> the server, wget will start all over again and will not resume.
>
> Here is a sample of the error:
> -------------------
> HTTP request sent, awaiting response... Read error (Connection timed
> out) in headers.
> ------------------

Try with the -S option, so you see the header you get, could be a 404 or
something that would cause a legal skip, there are servers that gives
quite bad errors at high load.


--

//Aho

On Mar 16, 3:24 am, "J.O. Aho" <u...@example.net> wrote:
> wdoe...@yahoo.com wrote:
> > Hi - I am having a problem with wget v1.10.2 when I run the command:
>
> > wget -t 100 -T 10 -i files.txt
>
> > On older versions (eg v1.6), this will keep retrying and will "resume"
> > the download from where it left off, EVERY time.
>
> From man page:
> --tries=number
> Set number of retries to number. Specify 0 or inf for infinite
> retrying. The default is to retry 20 times, with the
> exception of
> fatal errors like ``connection refused'' or ``not found'' (404),
> which are not retried.
>
> > On the latest wget version, this will do the retry properly if the
> > connection times-out in the middle of receiving bytes (read-error),
> > however, if there is a timeout when receiving the header response from
> > the server, wget will start all over again and will not resume.
>
> > Here is a sample of the error:
> > -------------------
> > HTTP request sent, awaiting response... Read error (Connection timed
> > out) in headers.
> > ------------------
>
> Try with the -S option, so you see the header you get, could be a 404 or
> something that would cause a legal skip, there are servers that gives
> quite bad errors at high load.
>
> --
>
> //Aho

Thanks for replying.
Yes, I noticed the instructions to which you refered.
I tried to use the swich "--retry-connrefused" and that didn't help.

I ended up solving it though. I used the -c switch.

The instructions say that the -c switch is used to "continue" a
download from a previous invocation of wget. The problem is that they
explicitly say that the -c option does not do anything for the
"current" invocation of wget. That is why I never tried it to begin-
with. After I thought about it for a while, I started to think that
wget is "essentially" starting a new download when it gets a fatal
error, and perhaps the instructions could be wrong. Indeed this is
true and the -c option does have affect in the current invocation.