I am getting a sprinkling of tasks which fail due to a file open error - just a handful across several systems - like maybe 4 in the last 36 hours across maybe 80 - 100 work units.

The error message is always like:

ERROR: Cannot open patchdock file: 1fAc_2j44.patchdock
ERROR:: Exit from: src/protocols/ProteinInterfaceDesign/read_patchdock.cc line: 101

Just want to make sure that my version of "gunzip" or whatever Rosetta calls is not the cause.

The zip support is all bundled into the BOINC Manager. So it shouldn't be any problem related to your version.

Is the problem file always the same name? Or always a patchdock extension?

Probably best to report this in the v2.14 thread (assuming that is the version involved with the task).

Were these failures rather immediately in the life of the task? Did reissued task fail same way on other machines? If you could post links to a few we can show you how to answer the above if you are not aware.

---

Rosetta Moderator: Mod.Sense

Thank you for the quick response. To try and answer your questions:

1. No, it is not always the same exact filename, though I do have a couple where the same filename was involved.

2. Yes, it is always the "patchdock" file

Here is a "cut-n-paste" of the file names which failed over the past few days. There have been 6 failures across 5 systems. Further, 3 of my servers did not see this error.

During this time frame I also processed several hundred work units without error so it is not a persistent nor a consistent error. I prepended the name of the server involved on each record just for my own <limited> sanity in case you wanted more information on any specific failure.

Popeye: ERROR: Cannot open patchdock file: 1fAc_2j44.patchdock
Proteus: ERROR: Cannot open patchdock file: 1fAc_2odh.patchdock
Neptune: ERROR: Cannot open patchdock file: 1fAc_2vg9.patchdock
Poseidon: ERROR: Cannot open patchdock file: 1fAc_2vg9.patchdock
Poseidon: ERROR: Cannot open patchdock file: 1fAc_2boo.patchdock
Sinbad: ERROR: Cannot open patchdock file: 1fAc_2j44.patchdock

3. Yes, when the failure occurs it happens just a few seconds into the task.

I will try and post to the v2.14 thread after absorbing some caffeine.