walrus_request(): wrote X bytes in 0 writes

[Dmitrii Zagorodnov]

On some installations, nc.log contains lines of the sort:

  walrus_request(): wrote X bytes in 0 writes

Where X is non-zero. In theory this should not happen as the two counters always move up together. Perhaps this is a sign of some concurrency bug (though the behavior seems to be consistent on the installations where it occurs). Further investigation is needed.

[Dilraj (dilrajm)]

Hello Dmitrii,

I have installed Eucalyptus on Ubuntu 9.10 Karmic (Ubuntu UEC) using the following setups:
1) cluster/walrus/cloud controllers on 1 machine and node on another machine
2) everything on the same machine

Each time I get the above-mentioned error:
"walrus_request(): wrote X bytes in 0 writes"

Somehow, the files that was supposed to be copied in admin, instances have not been copied from the tmp directory.

And of course, later on libvirt complains that it has not found an appropriate kernel, etc.

Here are my technical details:
- Intel Core 2 Duo
- Virtualisation enabled in BIOS
- kvm-ok script returns "KVM acceleration can be used"

Could you please indicate me how to proceed in this case?

Thanks in advance.

Kind regards,
Dilraj

[Andy Grimm (agrimm)]

This issue is now being tracked upstream at http://eucalyptus.atlassian.net/browse/EUCA-2660

Please watch that issue for further updates.