cannot post comments on trac #19984

[Dima Pasechnik]

trac server basically hangs on hitting "Submit Changes"

Is it just me?

[Volker Braun]

There is a request data read error in the log, this seems to be a networking error and not the fault of the trac server. I tried and could comment on the ticket just fine.

[Dima Pasechnik]

On Sunday, 31 January 2016 12:23:59 UTC, Volker Braun wrote:

There is a request data read error in the log, this seems to be a networking error and not the fault of the trac server. I tried and could comment on the ticket just fine.urther

I keep getting the same problem; a further symptom is that live preview of what I type in the comment gets stuck after first 30-40 characters of so...
(I even tried to connect via a vpn, but it makes no difference)


 

[Daniel Krenn]

On 2016-01-31 14:07, Dima Pasechnik wrote:
> On Sunday, 31 January 2016 12:23:59 UTC, Volker Braun wrote:
>
> There is a request data read error in the log, this seems to be a
> networking error and not the fault of the trac server. I tried and
> could comment on the ticket just fine.urther

No problem at the moment with #10519.

> I keep getting the same problem; a further symptom is that live preview
> of what I type in the comment gets stuck after first 30-40 characters of
> so...
> (I even tried to connect via a vpn, but it makes no difference)

I have this every now and then with trac. I have no idea whats wrong.
Sometimes it helps to split the posting up into several parts; sometimes
even clicking on "Preview" at the bottom helps.

Daniel

[Daniel Krenn]

On 2016-01-31 14:19, Daniel Krenn wrote:
> On 2016-01-31 14:07, Dima Pasechnik wrote:
>> On Sunday, 31 January 2016 12:23:59 UTC, Volker Braun wrote:
>>
>> There is a request data read error in the log, this seems to be a
>> networking error and not the fault of the trac server. I tried and
>> could comment on the ticket just fine.urther
>
> No problem at the moment with #10519.

Now, some troubles. I split it into two comments and it worked. (but
it's annoying...)

Daniel

[Dima Pasechnik]

I can comment just fine on some tickets, and just cannot on some others. 

It really does not look like a networking error, but rather some trac trouble...

[Volker Braun]

I've increased the mod_reqtimeout limits, if we are hitting a timeout on the trac side then it must be that. Though it could very well be a timeout on the network in front of trac...

[Dima Pasechnik]

On Monday, 1 February 2016 00:28:31 UTC, Volker Braun wrote:

I've increased the mod_reqtimeout limits, if we are hitting a timeout on the trac side then it must be that. Though it could very well be a timeout on the network in front of trac...

still the same sh*t. It hangs on the automatic previewing. Is it possible to shut it off (perhaps have a setting to shut it off)?

Otherwise, someone would have to develop a fix for 'git trac' to post comments...

[Volker Braun]

You should be able to post even if the preview hangs... possibly requires to press send again. Did you try that?

[Daniel Krenn]

On 2016-02-01 12:28, Volker Braun wrote:
> You should be able to post even if the preview hangs... possibly
> requires to press send again. Did you try that?

When I had the same problem, I've tried, but it did not work.

Daniel

[Volker Braun]

can somebody make a packet trace, eg. using wireshark?

[Dima Pasechnik]

On Monday, 1 February 2016 13:49:28 UTC, Volker Braun wrote:

can somebody make a packet trace, eg. using wireshark?

I've made such an attempt and attached is the result. Hopefully there are no passwords there... :-)

[Dima Pasechnik]

is it just skrew clock somewhere upstream that causes this?

[Volker Braun]

Something broke already before you pressed submit, the next-to-last preview already timed out.... Still looks more like a network issue, perhaps a wonky UW firewall?

[Dima Pasechnik]

On Monday, 1 February 2016 20:43:40 UTC, Volker Braun wrote:

Something broke already before you pressed submit, the next-to-last preview already timed out.... Still looks more like a network issue, perhaps a wonky UW firewall?

you appear to be right; I was able to leave http://trac.sagemath.org/ticket/19984?replyto=22#comment:24

using geom as socks_v5 proxy (see e.g. https://calomel.org/firefox_ssh_proxy.html on how to set it up)

And a no web-proxy attempt still hangs.

What can we do about it? Complain to UW, via William?

Move over to other hosting?

(ODK grant surely can get us a good enough GCE or other host...)

Dima

[Nathann Cohen]

Hello everybody,

Just reporting that the same happened to me today. I tried to post four times on #19451 (forgetting *every single time* to save my message before hitting 'submit'), and no luck so far. It waits, it waits, it waits, then gives up.

Nathann

[Jori Mäntysalo]

I had the same problem few days ago for a one day or so. Haven't happened
after that. I did not change any network setting. So we really have some
heisenbug at trac.

--
Jori Mäntysalo

[Dima Pasechnik]

As I noticed, using geom.math.washington.edu as socks_v5 proxy for
Firefox does seem to work. But this is of course not an easy solution,
in particular this means that every Sagemath dev would need an account
there...

[Eric Gourgoulhon]

Hi,

Just to tell that I also have exactly the same problem from time to time (last time, a few days ago). Sometimes, changing the browser (from Firefox to Chromium) helps, sometimes not...

Best wishes,

Eric.

[William Stein]

On Mon, Feb 1, 2016 at 3:50 PM, Dima Pasechnik <dim...@gmail.com> wrote:
>
>
> On Monday, 1 February 2016 20:43:40 UTC, Volker Braun wrote:
>>
>> Something broke already before you pressed submit, the next-to-last
>> preview already timed out.... Still looks more like a network issue, perhaps
>> a wonky UW firewall?
>
>
> you appear to be right; I was able to leave
> http://trac.sagemath.org/ticket/19984?replyto=22#comment:24
> using geom as socks_v5 proxy (see e.g.
> https://calomel.org/firefox_ssh_proxy.html on how to set it up)
>
> And a no web-proxy attempt still hangs.
>
> What can we do about it? Complain to UW, via William?
>
> Move over to other hosting?
> (ODK grant surely can get us a good enough GCE or other host...)

SageMath, Inc. is probably large enough to pay for this now.

Any thoughts about the costs of the compute resources that would be
needed to move trac.sagemath.org to GCE? If it's not too much, it
would be easy to setup a GCE project for that, billed to SageMath,
Inc., and add the people who would be willing to maintain the
trac.sagemath.org and wiki.sagemath.org to that project. That's how
sagecell is done right now (it costs about $100/month, several of us
are on the project and all can do anything, etc.).

Quick estimate: wiki+trac would cost about $110/month?

https://cloud.google.com/products/calculator/#id=7efdfcf7-1d16-41b7-84b2-f98c50558c10

I'm willing to pay for this if a couple of people will volunteer to be
the people to do the actual move and maintain the GCE project. (GCE
is really easy to use through a web interface, including ssh, etc.)

-- William

>
> Dima
>
>>
>>
>> On Monday, February 1, 2016 at 5:49:54 PM UTC+1, Dima Pasechnik wrote:
>>>
>>> is it just skrew clock somewhere upstream that causes this?

--
William (http://wstein.org)

[Dima Pasechnik]

I would like to volunteer, although I don't know much about trac and the git server

used there. And my time is sometimes very limited, 

in particular during teaching weeks, i.e. about 20 weeks a

year at the moment.

(although it's actually written in the ODK grant proposal that we'd like to look

at anonymised statistics of trac usage, so this gives me an extra incentive to

look into details of trac functioning)

Dima

[William Stein]

So far only you and Thierry have stepped up. If you send Thierry
(sa...@lma.metelu.net) your public key, he can make an account for you
on wiki.sagemath.org and trac.sagemath.org.

> --
> You received this message because you are subscribed to the Google Groups
> "sage-devel" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to sage-devel+...@googlegroups.com.
> To post to this group, send email to sage-...@googlegroups.com.
> Visit this group at https://groups.google.com/group/sage-devel.
> For more options, visit https://groups.google.com/d/optout.



--
William (http://wstein.org)

[Samuel Lelievre]

2016-02-04 03:53:43 UTC, William:

So far only you and Thierry have stepped up.   If you send Thierry
(sa...@lma.metelu.net) your public key, he can make an account for you
on wiki.sagemath.org and trac.sagemath.org.

I'm interested in joining too. I'll send Thierry my public key.
Samuel

[Dima Pasechnik]

I emailed to Thierry at the address you gave
two times, from two different email addresses, never got a reply.
Do you know any other means to get in touch with him?
Thanks,
Dima

[Dominique Laurain]

@Dima : did you try

https://wiki.sagemath.org/ThierryMonteil
=> professional page: http://monteil.perso.math.cnrs.fr/
and after reading it, you have University Paris 13 email