Closed vishnevskiy closed 9 years ago
First off, we are now tracking issues through https://tokutek.atlassian.net/browse/MX. This looks like https://tokutek.atlassian.net/browse/MX-1308, which I'm not sure what is happening. Can you attach a gdb to the old primary, linode-prod-mongodb-0002:27017 and email the output of `gdb --batch -ex "set pagination 0" -ex "thread apply all bt" $MONGOD_PID to the tokumx google group?
Before doing so, you'll need to install debug info. Assuming you are running 2.0 community from a tarball, please download and install https://s3.amazonaws.com/tokumx-2.0.0/tokumx-2.0.0-linux-x86_64-debuginfo.tar.gz in the same directory you have tokumx installed.
Sorry about that I will post in JIRA next time.
That said I already did an initial resync =(
Thanks. That said, you are the second to hit this, so there is likely something wrong. I believe the original primary is stalling underneath the covers. If you hit this again and can get gdb stack traces, that will help greatly. Thanks.
I was trying to debug why a cron wasn't updating the database only to find out that there are 2 masters.
This is the output of the most up to date master.
And this is from the other master.
This is the version installed on all boxes