I am no longer getting new work. The message makes some sense, because I am running Ubuntu 16.10. But I have been for the last several months without problems.

There is something going on.

---

It is now working again on that machine, and I have added two more that are downloading new work properly. So whatever it was is now fixed.

---

I am no longer getting new work. The message makes some sense, because I am running Ubuntu 16.10. But I have been for the last several months without problems.

There is something going on.

This has been around for some time without any Rosetta attention and is not "fixed" but the Rosetta operation has just stopped getting to the bug.

I "suspect" that the developer who added Android support introduced one or more bugs in the server work load scripts. I have seen the "Android" message when the job queue has been drained and my computer tries to get more work.

When this happens, I tend to ALSO see Rosetta back off 24 hours for the next contact.

My "workaround" has been to manually monitor Rosetta and "UPDATE" the project from the BOINC ... in other words ... tell BOINC to try to talk to Rosetta again. The 24 hour back off is refreshed to its several minute normal ... unless the workload queue is still depleted.

Thanks. I had never run across it before.

---

My "workaround" has been to manually monitor Rosetta and "UPDATE" the project from the BOINC ... in other words ... tell BOINC to try to talk to Rosetta again. The 24 hour back off is refreshed to its several minute normal ... unless the workload queue is still depleted.

Now also on Ralph...
But why 24h? It's a boinc server default???

---

I'm seeing the same thing on Ubuntu 16.04: a project backoff of 24 hours accompanied by the message "Rosetta Mini for Android is not available for your type of computer." when attempting to get new tasks via Update.

Before the recent hiccups this machine crunched Rosetta without any problems.

Indeed, it is as if at the moment, the only work units being created are for the android phones and not for workstations.

For now, worldgrid is very happy <smile>

I'm seeing the same thing on Ubuntu 16.04: a project backoff of 24 hours accompanied by the message "Rosetta Mini for Android is not available for your type of computer." when attempting to get new tasks via Update.

Before the recent hiccups this machine crunched Rosetta without any problems.

I'm getting this error myself. Server keeps trying to send Android work to my Windows 7 machine.

My "workaround" has been to manually monitor Rosetta and "UPDATE" the project from the BOINC ... in other words ... tell BOINC to try to talk to Rosetta again. The 24 hour back off is refreshed to its several minute normal ... unless the workload queue is still depleted.

Now also on Ralph...
But why 24h? It's a boinc server default???

I suspect that the 24 hour back off is just an uninitialized variable being generated in the code under error conditions. I suspect there is subsequent code that checks for very large values and limits it to 24 hours. 24 is just a constant that someone chose.

When you debug code, verifying that the working path is OK is easy. Verifying that all the error paths are properly working is MUCH harder.

In this case, I suspect that the Rosetta developer who added the code to handle the Android case ... just introduced a number of bugs. It is also possible that the bug is older and in a different area but the uninitialized variable "blames" Android which is more visible.

Think about which Rosetta developer would feel the responsibility to clean up Android guy's bugs? 8-|

Think about which Rosetta developer would feel the responsibility to clean up Android guy's bugs? 8-|

No one, at this time, is resolving the problem...

---

Again, on Ralph

---