-
Notifications
You must be signed in to change notification settings - Fork 492
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
NAS-128893 / 24.10 / Remove md/swap usage #13863
Conversation
We need to remove stat reporting for swap and a few tiny things
Readded Ref
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Do we recall if we used to create persistent md mirrors because in that case they would still be around and we might probably still need to remove them and then format disks ?
@@ -66,8 +66,6 @@ async def replace(self, job, oid, options): | |||
# If the disk we are replacing is still available, remove it from swap as well | |||
swap_disks.append(from_disk) |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Remove this line (and swap_disks
variable altogether), but keep from_disk
as it is being wiped later
We've used to format boot device to use with only BIOS/EFI boot. We don't do this anymore, but we have to keep supporting formatting boot devices like this in order to be able to create mirrors for existing boot pools. Your changes are correct. |
I found this PR, which seems to be making those persistent swap mirrors. I'm still going through it all to figure out how I'd check for old persistent mirrors and do said removing of them (if I need to). |
@aiden3c you can use the same command as in the PR to create persistent mirror and then try formatting disks to see if it works with your changes. If it does not, then that means we still need to break/destroy that mirror before actually trying to format the disks involved as that mirror would come up. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
LGTM!
@aiden3c so the path of least resistance might be to run it manually yourself without going to a scale version which actually had those changes. If you run into issues with that, then yes - the next best thing would be to go to a scale version where we were creating these mirrors. Motivation behind this is that users who were on this scale version and had these mirrors setup, when they come to a version with your changes - they would still have those mirrors and format would likely fail in that case which we want to be sure doesn't happen. |
This PR has been merged and conversations have been locked. |
We're removing swap entirely from Middleware, refer to this parent Jira ticket for more information.
This PR has removed all instances of swap minus the four following areas:
10-truenas.conf
sets kernel swappiness. Unsure the correct value to set here, or if the line should entirely be removed. I'm leaning remove the line.htoprc
sets swap as a statistic to include. Modifying this doesn't seem to make any change on a VM.replace_disk.py
withinreplace
. I was going to just remove the entire section, but we usefrom_disk
at a later point. Wanted to make sure that someone reviews if both that section and the other usage offrom_disk
(in the try/except/else) can be safely deleted.boot.py
withinlegacy_schema
. I'm not entirely sure what is being done here (compared to get_boot_type alone).