Thread: gitmaster down?
Hi, Right now I can't reach gitmaster via ipv4 or ipv6. I tried from different networks, one in Europe one in the US. A bit before that gitmaster worked, but was extremely slow... Greetings, Andres Freund
Andres Freund <andres@anarazel.de> writes:
> Right now I can't reach gitmaster via ipv4 or ipv6. I tried from different
> networks, one in Europe one in the US.
> A bit before that gitmaster worked, but was extremely slow...
FWIW, I successfully did "git pull" just a few minutes ago
(Mar 24 19:33 -04:00, per file timestamps). However, I then
disconnected the ssh connection I'd had open since the
morning, and now I can't reconnect:
ssh: connect to host gitmaster.postgresql.org port 22: Operation timed out
regards, tom lane
On 3/24/25 16:38, Andres Freund wrote: > Hi, > > Right now I can't reach gitmaster via ipv4 or ipv6. I tried from different > networks, one in Europe one in the US. My money is on either: a) Return of Cisco router/DOS issue. b) Fallout of fix to above. > > A bit before that gitmaster worked, but was extremely slow... > > Greetings, > > Andres Freund > > -- Adrian Klaver adrian.klaver@aklaver.com
> On Mar 24, 2025, at 17:42, Adrian Klaver <adrian.klaver@aklaver.com> wrote: > a) Return of Cisco router/DOS issue. > > b) Fallout of fix to above. And the answer is: (a). The firewall once again went down. It's been rebooted, and service should be restored shortly. (The firewall is going to be replaced relatively soon, and that should put an end to these shenanigans.)
On 3/24/25 17:43, Christophe Pettus wrote: > > >> On Mar 24, 2025, at 17:42, Adrian Klaver <adrian.klaver@aklaver.com> wrote: >> a) Return of Cisco router/DOS issue. >> >> b) Fallout of fix to above. > > And the answer is: (a). The firewall once again went down. It's been rebooted, and service should be restored shortly. (The firewall is going to be replaced relatively soon, and that should put an end to these shenanigans.) I don't know if it related but if I go here: https://git.postgresql.org/gitweb/?p=postgresql.git;a=tree;f=src/common;h=38a89f2dd58609282d30c662c970d239a550b0d3;hb=HEAD and click on a file I get: a) In Firefox a blank screen. b) In Chrome this: This page isn’t working If the problem continues, contact the site owner. HTTP ERROR 429 -- Adrian Klaver adrian.klaver@aklaver.com
On Tue, Mar 25, 2025 at 5:14 PM Adrian Klaver <adrian.klaver@aklaver.com> wrote:
On 3/24/25 17:43, Christophe Pettus wrote:
>
>
>> On Mar 24, 2025, at 17:42, Adrian Klaver <adrian.klaver@aklaver.com> wrote:
>> a) Return of Cisco router/DOS issue.
>>
>> b) Fallout of fix to above.
>
> And the answer is: (a). The firewall once again went down. It's been rebooted, and service should be restored shortly. (The firewall is going to be replaced relatively soon, and that should put an end to these shenanigans.)
I don't know if it related but if I go here:
https://git.postgresql.org/gitweb/?p=postgresql.git;a=tree;f=src/common;h=38a89f2dd58609282d30c662c970d239a550b0d3;hb=HEAD
and click on a file I get:
a) In Firefox a blank screen.
b) In Chrome this:
This page isn’t working
If the problem continues, contact the site owner.
HTTP ERROR 429
That's unfortunately correct. We've had to block a few operations in gitweb due to being DDOSed by "ai bots"... Binary operations and "blame" are off due to gitweb (and for blame, also git) being incredibly slow at it.
On 3/25/25 09:25, Magnus Hagander wrote: > > That's unfortunately correct. We've had to block a few operations in > gitweb due to being DDOSed by "ai bots"... Binary operations and "blame" > are off due to gitweb (and for blame, also git) being incredibly slow at it. Would it be difficult to add a message here: https://www.postgresql.org/developer/coding/ that clues folks in? > > -- > Magnus Hagander > Me: https://www.hagander.net/ <http://www.hagander.net/> > Work: https://www.redpill-linpro.com/ <http://www.redpill-linpro.com/> -- Adrian Klaver adrian.klaver@aklaver.com
On Tue, Mar 25, 2025 at 5:31 PM Adrian Klaver <adrian.klaver@aklaver.com> wrote:
On 3/25/25 09:25, Magnus Hagander wrote:
>
> That's unfortunately correct. We've had to block a few operations in
> gitweb due to being DDOSed by "ai bots"... Binary operations and "blame"
> are off due to gitweb (and for blame, also git) being incredibly slow at it.
Would it be difficult to add a message here:
https://www.postgresql.org/developer/coding/
that clues folks in?
Maybe better to put that on the frontpage of git.postgresql.org itself?
We could also add a link to the github mirror to the coding page above perhaps?