Protocol file errors

Discussions
1

The problem
I have a problem with saving changes in a protocol that worked normally before. It started when I was doing some work in Brainstorm (deleting some source files), and suddenly started to receive an error message when trying to save the protocol.

  • When running a process in BS, sometimes I get the error message that the file could not be saved, because either (1) the disk is full, (2) disconnected, (3) read-only.
  • In the Matlab console I get a “Cannot read protocol file” error message

What I already tried
My problems seem similar to those from another forum post (Corrupted protocol). I tried the suggested solutions to no avail.

  1. Detach the database, move it, and reload it from new location.
  2. Delete the protocol file and reload database

Technical information

  • OS: Linux CentOS 7, kernel version 3.10.0-693.5.2.el7.x86_64;
  • Matlab Version: 2014b;
  • Brainstorm version: build 160905
  • brainstorm db is stored on a remote storage, which is mounted on the CentOS using NFS with the following options: rw,relatime,vers=4.1,rsize=1048576,wsize=1048576,namlen=255,hard,proto=tcp,port=0,timeo=600,retrans=2,sec=sys

I hope someone can help me fix this.

Best,
Eelke

2

First, the version of Brainstorm you are using is old and not supported anymore (september 2016). Please update to today’s version before trying to solve your issues.

The instructions you got from the post you mentioned is now incorrect, the protocol file is now most likely located in protocol_folder/data/protocol.mat. Is this the file you have delete after detaching the protocol?

A few additional debugging questions:

  1. Does this issue happen randomly? Or is it strictly reproducible?
  2. Is it happening for all the protocols or only on one of them?
  3. If you duplicate the protocol (copy-paste the protocol folder and load the copied folder with the menu File > Load protocol > Load from folder), do you get the same issue?
  4. If you copy the protocol on a local drive, does it work any better?
  5. Have you tried from a different computer with the same NFS mount?
3

Best Francois,

Thanks for your fast reaction!

The problem seems to have resolved itself. The database directory (containing the protocol) was moved to another location on the server this weekend. After reloading the database, everything works as it should again. Somewhat strange, because I tried moving and reloading it before, and that didn’t solved anything…

So I’m happy for now. I’ll update this post if the error returns or when I find out what was causing it.

Thanks!
Eelke de Vries