Closed lnsyyj closed 2 years ago
How can I reproduce it?
My steps are:
1. Create a pool
2. Create inode 200G
3. Via vitastor-nbd map inode
4. Create a file system mkfs.xfs /dev/nbd0
5. Mount the file system mount /dev/nbd0 /root/nbd0/
6. Then I copied a lot of code to the /root/nbd0/ directory
cp -r github/ /root/nbd0/github, cp -r github/ /root/nbd0/github2, cp -r github/ /root/nbd0/github3...
So it will reappear
Hi, I recently fixed a very simple "splitbrain" issue (commit aa436027c8aa1a788a37f12e118fe46865cb70f0) which I missed in previous versions. It could easily manifest as data corruption if you had 2 osds and if you were stopping them one after another to test failover :-). May it be similar to your case?
Hi @vitalif , The problem is solved in version 0.6.10, I think it can be closed.
Cool
When I use nbd to mount the vitastor image and read and write, the following error occurs