-
Notifications
You must be signed in to change notification settings - Fork 22
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
ERROR: You must have read and write access in the current directory #12
Comments
Interesting. The very first line of the output there seems to be where it started going wrong:
That's a very strange error for it to output, but then keep on going. Oh... I'm guessing that's probably not the output from the very first run (?), so it probably already has the It looks like it's successfully creating the But it looks like the whole process is dying shortly afterwards, prior to the At a minimum, we should probably add some precautionary checks to the start of the process to make sure there's no existing Also, it probably wouldn't hurt to add some more verbose progress logging so that knowing what's going on is easier. I'll make some updates now, and push a |
With the mentions of I've not tried that out yet, but if the same problem happens with Matrix databases created for a new setup then that would give me something to experiment with. 😄 |
Hmmm, there seem to be several backend server possibilities: https://matrix.org/ecosystem/servers/ If this is the correct Matrix thing, please let me know which one I should try out in order to replicate the problem locally. 😄 |
Yes this is for a Matrix server that we run for our community. It is setup and configured using this ansible playbook (https://github.com/spantaleev/matrix-docker-ansible-deploy). We need to upgrade the postgres database from 14 to 15 and the documentation is to export the database and then re-import which with a database that is 300 to 400 GB takes a long time so we are looking for a better way to upgrade the database. |
Yeah I would definitely be up for testing changes to see if they work. |
The server that we are using is synapse |
Thanks. I'm adjusting things with the extra logging at the moment. Need to run the test harness against that (shortly), so probably have something you can try in about 20-30 mins. 😄 |
k. I've just pushed a new docker image with the plain It's super likely to break in the exact same way as before, but at least now it should output enough info to narrow down WHERE and potentially WHY it's breaking. 😄 |
Oh, I wonder if the The existing logging doesn't report that current directory. I'll fix that now, 1 minute... |
I've just pushed a new The failure really could be that simple to fix. 😄 |
I ran with the dev image and it did make it further but it then failed again.
|
Thanks, that's super helpful. 😄 Two messages in the log are standing out to me:
and:
The first one is clearly not fatal, but will need investigating anyway for safety. The second one clearly is. It's probably a result of the initdb arguments you showed in your original message:
I've not personally come across a situation before where UTF8 is used with the C locale. There's nothing really wrong with doing that however. I'll update the scripting to use that environment variable if present, which should (in theory) make it more robust and also work for this situation. 😄 |
Btw, after that above failure your entire (working) set of PostgreSQL 14 database files is in the You should be able to nuke the |
I'm substantially refactoring the upgrade scripting at the same time, so it'll probably take a few more hours before there's something you can test. 😉 |
@darkdecoy There's a new This one has about an 80% chance of working completely successfully for you. 😄 |
This time it did the upgrade and started the database server successfully. Going to do some further tests to ensure that the matrix server continues to function but what I have seen so far this is a much faster upgrade path than exporting the data and then reporting which took like 2 days. |
Awesome! 😄 Out of curiosity, how long did it take for this pgautoupgrade approach to run (prior to the database start)? |
Started systemd service at Aug 21 22:14:47 |
19 seconds. Cool, that's a pretty fast upgrade compared to 2 days. 😁 |
When trying to move from postgres 14.7-alpine to pgautoupgrade/pgautoupgrade 15-alpine3.8, I keep getting the error You must have read and write access in the current directory.
Log Output:
docker run command:
Docker Environment File:
The text was updated successfully, but these errors were encountered: