Why are you missing one satellite?
My node migration is not doing well.
In order to minimize node downtime, I decided to run ārync --deleteā while the node is running. Well, deleting files takes forever, itās been almost a full day to delete files from a single satellite. When it finishes deleting, it will copy all the deleted files to the trash directory. Rsync is loosing terrain instead of gainingā¦
Really really bad timing for a migrationā¦
Because the satellite northern Europe lies dead weight, - only for storage.
Look on the bright side - this is the best time to shrink the size of the node.
yeahā¦ but I wasnāt trying to shrink it, just migrate itā¦
And for getting new nodes vetted
Out of my ttl 10TB of stored data, 1 TB moved to the trash in the past 3 days.
I can accept that this is a centralized action to get rid of some test data, but why it is done like this?
I mean this huge deletion is a kind of stress test?
Otherwise it would be much better from node operatiorsā point of view to do it in smaller steps.
Iād say itās been done in small stepsā¦ a bit every month for the last few monthsā¦
What does āttlā mean?
Some data expires, depending on how it was uploaded originally. Weāve been working with a lot of companies for a while now. Sometimes a company may want to test their operation against Storj. For instance, a company might have 500tb of data that they want to upload every six months, then erase it, and then repeat an upload of another 500tb. Or they may want to upload 1tb in 300,000 segments. They may want to test the speed of Storj, its ability to handle large datasets, its retrieval, accounting of segments, and various other things. At some point, when these operations end, the customer may purge this data or it may expire. These are just examples. Large customers donāt always just sign up and start using a product like this. They work with the sales team and executives, discuss their needs, and we work with them to test and migrate what they want done. Those kinds of discussions can take a long time. Tests at different scales can take a long time. Itās normal for this kind of business.
The data is anonymous, so it could be any kind of situation over any period of time.
On the plus side, old data tends not to be pulled on as much as new data. So removal of old data on your nodes provides room for new data which may give you more egress.
Wellā¦ I saw weekly deletions of 50-100GB from the same node earlier, now it is 700GB at once.
ttl means total.
Thank you for this explanation!
The only thing that surprise me is the large amount of deleted data. My nodes are storing around 10TB in total, now 1TB was deleted within 2 days. This is about 10%. I see similar rate on the uploaded images.
Considering that the total stored customer data is 17PB, around 1700 TB was deleted nowā¦
It doesnāt scale this way. You can see from the official statistics that about 400TB have been deletedā¦
OK, so 1TB for me, another 1,1 TB for naxbc, another 1,1 TB for AiS1972, just from this topicā¦ 3 of us covered a bit more than 3 out of 400TB. Considering the 16k nodes (this was the nr of vetted nodes 6 month ago) this seems quite uneven data distribution for me.
And I still have the data in the trash, recoverable. Is it counted as stored customer data or not? If not counted and they are 100% sure that they donāt want to recover it, why is it sitting in the trashā¦?
At the end of the day, Iām happy that some space free up, as new data gives me more egress ā $$$
Iāve got 1.32TB in Trash, but I know my node is larger than most.
Everyone whoās posted has a larger than average node. Some nodes are full and have been for years. Some nodes are 500GB. Some are 1TB. Some are in Australia or South America.
This is the point of decentralised storage, all nodes are different hardware, in different places.
This has been answered before. Itās not paid unless a recovery takes place. Itās in Trash precisely so Storj can recover it, if an error occurs.
Iād prefer to have both the old and new data . The more there is the more I get paid!
Two of my nodes, behind the same IP:
Add ~600GB trash from ~10TB storage from me (from 2020 node).
for me it is a bit above 1TB in trash across all my nodes.
It is more than iāve ever seen, but my nodes keep on running.
The thing concerning me is that i now see overusage on the dashboard.
The overusage should be self corrected after a while.