Perhaps you mixed ports?
This is mean, that you have had a newer version than the current upstream. The downgrade is not supported. You now need to run the newest version on these nodes.
can you tell me how i run the newest version on these nodes?
No, you should tell me, how you did that and bypassed the storagenode-updater.
Right now the easiest way is to upgrade to a new version, how you hack it before, or delete all databases.
i honestly have no clue.
can i run a version of this command that is not breaking things but still gets the newest verison downloaded?
sudo docker run --rm -e SETUP=“true”
–user $(id -u):$(id -g)
–mount type=bind,source=“/home/andreashg/Desktop/Identity/storagenode4”,destination=/app/identity
–mount type=bind,source=“/mnt/sdc1/StorjData”,destination=/app/config
–name storagenode4 storjlabs/storagenode:latest
it seems the nodes are on 1.105.4
and not 1.108:
so what i need to do is fix the error with database mix match somehow? and i can do this by updating what exactly?
i see most nodes is on 1.105
but one of them is actually 1.108
no idea how or why
Yes, you can.
First - do not execute the SETUP step ever, it will kill your node eventually. It must be executed exactly only one time, when you generated a new identity and never again. Any next may (and will) destroy your node.
because it’s eligible (the update cursor on https://version.storj.io instructs the node, can it be updated now or not).
Two failed seems somehow received the new version (1.106.x or later) and now downgraded to 1.105.x
Here are only two solutions:
- Forcibly upgrade them to 1.108.x and do not restart them, until the release is finished (then the minimum version would be bumped to 1.108.x, so no silent downgrade would be possible).
- Delete databases (and lose stat) and restart the node.
deleting the database for the nodes seems to work. atleast for the one i tried now.
but all the data in the dashboard will eventually return right? after file walkers and such?
Not everything, only information about used space and past payments from trusted satellites; data for untrusted satellites will not be restored. The bandwidth usage will be shown incorrectly, including the estimations on the Payout information page for the current period.
1 Like
Since you are using device names for mountpoints - like sdf- I would expect that you also use device names for mount devices in fstab?
these can often change during reboots. It mostly tied into spinup time of the individual drives etc.
In this case it sounds like what you had as drive sdf before reboot is now actually mounted at another path. Like this you THINK it’s the same node based on mount path, when in reality it’s another nodes data. Then you pair it with a wrong identity.
Suggestion:
Check your fstab and make sure you mount based on by-id and it would also make sense to use sensible names for mountpoints - like 1,2,3…
Just a thought.
4 Likes
thank mike - this post is a bit old and not relevant anymore- but your analysis was 100% correct and it was what fixed it back then 
1 Like
LOL
I didn’t notice the date 
It’s a problem with discourse. It clobbers together “new” and “unread” topics in a single pile:
Many times I’ve wasted time writing a comment only to delete it without posting because i would realize im responding to an ancient post while proofreading.
Well, maybe it a not a problem, perhaps it’s by design to stimulate engagement. Either way, very annoying experience
5 Likes
Yes.. don’t we all know it :-/
This was actually on my phone - so it’s even harder to spot these things. And I thought i had proofread it - but reading my post now, it certainly looks like a phone post 
1 Like
I don’t have this problem, Discourse looks like this for me:
The red line is pretty clear indicator to stop at this point. Maybe it’s a matter of some configuration switches?
I have this line on the computer as well - but i’m mostly browsing through this forum on my phone, not sure if it’s as visible.
Come to think of it, i believe i clicked on top topics and ended up on this thread. During this - i just didn’t check the date of the latest post. My bad i know