Open vojtechsimetka opened 2 years ago
Just run into this. Not sure it happens only on "reuploaded" data... but mostly on slightly larger files... 15+MB... i can upload small files repeatedly without issues
You might want to look at, and compare, the batch details after the first and second uploads to see what is changing in the batch. I presume that you're not locally pinning the uploads? It may only happen if the gc happens to flush out chunks of the upload that aren't in the uploading neighborhood's reserve. This would be much more likely to happen on the mainnet than on the testnet due to the volume of stored data and possible churn of the cache. https://docs.ethswarm.org/debug-api/#tag/Postage-Stamps/paths/~1stamps~1%7Bid%7D~1buckets/get
You might want to look at, and compare, the batch details after the first and second uploads to see what is changing in the batch. I presume that you're not locally pinning the uploads? It may only happen if the gc happens to flush out chunks of the upload that aren't in the uploading neighborhood's reserve. This would be much more likely to happen on the mainnet than on the testnet due to the volume of stored data and possible churn of the cache. https://docs.ethswarm.org/debug-api/#tag/Postage-Stamps/paths/~1stamps~1%7Bid%7D~1buckets/get
My thought were more local because it happens only with upload through gateway-proxy while direct upload to bee works.
My thought were more local because it happens only with upload through gateway-proxy while direct upload to bee works.
Ah, that wasn't clear in the issue description. Unless the gateway is adding more APIs that your direct-to-bee testing uses, then this would truly be a mystery.
Sometimes, upload fails due to stamp being over issued. This happens when same data are reuploaded.