I just noticed that I've been experiencing the same issue on select machines for ... a couple days(?) I haven't read this complete thread, but was planning to see what might be common and why there's a difference.

BTW ... the message log on the affected machines reads the same as the other posts. It should be noted that several of my affected units are in differenct locations ... on differenct networks ... different ISP's, etc. None have actually experienced a problem with normal network functionality.

The standard boinc message log (without additional debug levels activated) appears to indicate there's a problem making a connection to a server that is providing the work units. I also had a server fail to download the core files after successfully attaching to WCG.

I've noticed a couple people trying different fixes, but no silver bullets.

Did I miss a solution somewhere along the line?

Thanks ... I'll go do some more reading. I can setup a tunnel to each of the remote networks and get to the machines pretty easily.

Just a matter of time.... never seems to be enough. :-)

Any speculation as to a root cause or long-term solution?

Best regards,

T

Additionally, machines are Erroring out because they can't download the files, causing unnecessary repairs to be issued:

<core_client_version>6.10.17</core_client_version>
<![CDATA[
<message>
WU download error: couldn't get input files:
<file_xfer_error>
<file_name>hcmd2.1C1U_H.clustersOccur.pdb.gzb</file_name>
<error_code>-200</error_code>
</file_xfer_error>
</message>

I think it's about time someone reminded the techs that we provide our machines for useful work, not just to waste tons of bandwidth (which, you know, some of us have to pay for) and maybe they should at least acknowledge the problem and inform all users about it.

It's not mine. That's one that died that I received a repair job for. I used the /etc/hosts "fix" after my machines did nothing except suck bandwidth for over a day. Those who don't happen to look at their logs are in a far worse situation. ... And perhaps end up aborting the transfers....

<core_client_version>6.10.56</core_client_version>
<![CDATA[
<message>
WU download error: couldn't get input files:
<file_xfer_error>
<file_name>hcmd2.2ZGV_A.clustersOccur.pdb.gzb</file_name>
<error_code>-197</error_code>
<error_message>user requested transfer abort</error_message>
</file_xfer_error>
</message>

I've started to switch back to having normal /etc/hosts files and, so far, downloads are still working.