I would just remove the progress indicator altogether. It looks like it is confusing indeed, and all that matters is that the process was started, so a yes/no flag “exiting” is enough. The rest, as @arrogantrabbit says.
2 Likes
It seems that after GE finishes, there is no confirmation of success or a confirmation ID.
It just says:
No graceful exit in progress.
It’s like it never started.
Then it perhaps even not started?
Do you have old logs? They should have something related to GE.
Also, if you did remove the databases, it perhaps lost the GE request.
I didn’t remove dbs, there is nothing in the logs because I don’t use lazzy mode, and with that turned off, nothing is logged about walkers, except startup piece scan.
I regularly checked the GE status and it showed that is on with the usual Success-No, Receipt-, etc.
By now it should be finished, because I started more than 30 days ago.
The only thing that I did was to move dbs from boot drive to storage drive. I triple checked to not let anything behind.
I’m thinking to delete all dbs and the piece-expiration directory, and run a startup piece scan.
Than try to GE Saltlake again to see what happens.
I would interested in the results as well. Usually if the GE is finished, you should get a receipt or node would be DQ.
I do not expect to have a third state.
I stopped and rm the node, deleted dbs and pieces_expiration dir, I restarted with startup piece scan on, finished the scan and the same message: No graceful exit in progress. The ingress was zero, though.
I started again the GE process on the same satellite (SL), and now it shows the usual:
Domain Name Node ID Percent Complete Successful Completion Receipt
saltlake.tardigrade.io:7777 xxx 0.00% N N/A
So, the satellite knew that the node GE from it, but the node didn’t know this.
How can this info be lost?
I only moved dbs to storage drive and did a pragma check and vacuum.
What happens now? I wait again 30 days for the GE to finish and the data be sent to trash?
What about the satellite? It sees another GE from the same node…
What about my collateral? How can I check if the held back was returned or will be returned?
Can someone look into it? I can send the node ID.
EDIT:
After writing this, I checked again the status and the GE is complete:
Domain Name Node ID Percent Complete Successful Completion Receipt
saltlake.tardigrade.io:7777 xxx 100.00% Y xxx
WOW!
1 Like
It consist of two parts: in the satellite DB and in the node DB. I believe it’s not backported, if it’s missing in the node’s database.
It should be included to the nearest payout. You may check that on your dashboard, when you would have a link to TX in the Payout history. You also may see how much it was calculated for the particular satellite, either in the Payout history or using a Custom range on the Payout information page.
A post was split to a new topic: Trash is not updated after the Graceful Exit is completed
Hi all,
I started graceful exit over 30 days ago - and its still going, still egressing lots of data. I get no progress reported anywhere based on the guide, until completion, so I am just wondering - what do you do if it goes on more than 30 days?
Many thanks
CC
Then you are done. You can shut down the node, and follow the rest of the documentation.
Cheers dude - yeah I get that, and have already check the storagenode exit-status.
Thing is, its been >30days so far, and its not complete
From the documentation I already read that:
“After the graceful exit period (currently 30 days) is complete, your node will no longer be in the graceful exit stage.”
For me this is not the case - its still busily sending data. I dont want to stop the node, but I need to pull the plug soon, and I have done everything asked.
CC
Restart the node and restart the GE. Sometimes if you restart the node during GE, it can loose the stats, but as soon as you try to restart the GE again, it will update its status from the satellites. When the GE is done, you will get the receipt.
1 Like
What am I hinting at — it has been over 30 day. Regardless of what status shows. Shut down the node. Open support ticket and ask for held amount if you want it. You fulfilled the requirements. You already did way more than most. Most people would “killall storagenode && rm-rf /mnt/storagnode”. Graceful exist is economically unjustifiable.
There will always be egress — node is online and is serving repairs.
1 Like
This is a wrong advise. They should check is GE even started at all. Because right now it sounds like it’s not. Maybe they didn’t check, that it’s started or ignored a error when they called it. Or called on another node, if they use a host networking and multiple nodes but didn’t provide a different private address of the node in the command.
@Climbingkid please provide an output of the storagenode exit-status command
So GE process definitely started, and I watched as all incoming traffic stopped.
I cannot copy / paste text from TN shell for some reason, so excuse the screenshot of exit status, I had to point at my config and identity to avoid errors.
Thanks
CC
Did you followed my instructions? Nothing bad happens if you restart the GE on a node that already has GE in progress. It will only sync it’s status with sats. Dosen’t restart at the begining.
From your screenshot you are in one of the following situations:
- your node restarted and gone “out of sync” with sats.
- you started GE less than 30 daya ago.
I speak from expirience. I GE with the old way and the new way, so I encountered all situations.
2 Likes
Left it longer - and it has just finished. Data is deleting (slowly) 
Thanks all
CC
1 Like