Pages: [1]
shanen
BAM!ID: 215770
Joined: 2017-03-06
Posts: 5
Credits: 0
World-rank: 0

2017-03-09 20:12:40
last modified: 2017-03-09 20:14:44

By using the hosts file approach I was able to visit the entry server for the Rosetta@home project, and it was running fine. I even posted on their internal forums.

However, Rosetta@home remains inaccessible today. It should take only a few minutes to fix a DNS problem, not several days. Yes, then there will be some propagation delay, but this has gone pretty far beyond the point of being ridiculous. It has reached the point where you have to wonder about the fundamental technical competence of the people running the project. That even casts doubt on their results. If they can't figure out DNS, are their MUCH more complicated algorithms and programs sound? In my own experience, the hardest problems often involve cleaning up problems with the data, but DNS problems should be much easier to clean up... Unfortunately, this is NOT the first time I've wondered such things. On the one hand, I've been involved in a lot of these projects over the years, going back to the original seti@home project where I was in the top 1% of contributors (and where I even suggested implementing some of the features of BOINC before usenet died) and I have earned a lot of 'points' from rosetta@home, but on the other hand, maybe it's time to look for a new project and fish for recommendations?

There is the obvious workaround. If someone has a list of all of the required hosts and their current IP addresses (possibly with some aliases, too), then I could temporarily patch all of the hosts files on the various computers. Based on yesterday's test, that would probably work, but...

In terms of finding a new project, I can say that the #1 annoyance of EVERY project that I've used under BOINC was the deadlines. Apparently the BOINC client doesn't provide any scheduling help for the projects, so they can't assess how much work to download and what the deadlines should reasonably be set to. *sigh* :-<

P.S. Perhaps a bug in this system? It says to add my sig, but nothing appears?
lohphat
BAM!ID: 201174
Joined: 2016-02-11
Posts: 16
Credits: 642,680,033
World-rank: 1,868

2017-03-09 23:52:59

I contacted the University of Washington admin listed in the whois database and got this response:

----

We (bakerlab.org) had a domain verification lapse, and our registrar (dotster.com) and ICANN turned off DNS for bakerlab.org. We started going through the steps to getting it verified again Monday afternoon. What should have been a 10 minute procedure is now stretching into 4 days. I promise you that we're doing everything we can to get dotster and ICANN to resolve this.

The registrations for these domains are paid up through 2020. This is the result of the implementation of a fairly new and separate security measure. All our name servers are working, and if you point your machine's nameserver to one of ours, you could probably continue with you boinc projects. However I don't recommend that. I hope this problem will be solved soon, but it's out of our hands and I've hoped for a quick solution for for 3 days now.

I'm sorry for your inconvenience, and I really appreciate the contributions you and the many thousands of others have made to Rosetta@home

Sincerely,
Darwin
lohphat
BAM!ID: 201174
Joined: 2016-02-11
Posts: 16
Credits: 642,680,033
World-rank: 1,868

2017-03-10 01:04:18

shanen wrote:

By using the hosts file approach I was able to visit the entry server for the Rosetta@home project, and it was running fine. I even posted on their internal forums.

However, Rosetta@home remains inaccessible today. It should take only a few minutes to fix a DNS problem, not several days.


The problem is that the DNS problem lies at the registrar level, not with Bakerlab/UW -- it's out of their hands until the bureaucracy responds.
lohphat
BAM!ID: 201174
Joined: 2016-02-11
Posts: 16
Credits: 642,680,033
World-rank: 1,868

2017-03-10 15:55:37

UPDATE from the UW admin:

The DNS issue has been resolved, however it may take some time for your local DNS infrastructure to timeout the lookup failure and then resolve properly.

DNS servers save bandwidth by caching lookups AND failed lookups -- it's just a matter of hours before things work for your client automatically.
Pages: [1]

Index :: News :: Rosetta@home STILL down. Time for a new project?
Reason: