On Sun Jul 17, 2022 at 16:43 CET, 'Sebastien Binet' via Perkeep wrote:

[...]

> > is the only way to create a new replicate to scp-copy the blobs?
> > (or implement the correct blob-order enumeration? I could try my hand at
> > it, if pointed a bit at the correct way)
>
> I have tried the ssh-copy avenue.
> AFAICT, I didn't get the packing errors anymore but for some reason,
> starting perkeepd on the remote-storage always failed with:
>
> blobpacked: 0 large blobs found in index, 44723 missing from index
> sample missing large blob: sha224-0001b1...xxx
> sample missing large blob: sha224-0001bf...xxx
> sample missing large blob: sha224-000576...xxx
> sample missing large blob: sha224-000a2b...xxx
> sample missing large blob: sha224-000d89...xxx
> sample missing large blob: sha224-001086...xxx
> sample missing large blob: sha224-0011d3...xxx
> sample missing large blob: sha224-001399...xxx
> sample missing large blob: sha224-0014ca...xxx
> sample missing large blob: sha224-00153e...xxx
> Error: 44723 large blobs missing from index. Please re-start in recovery
> mode with -recovery=1
>
> even after restarting with -recovery=1 and/or -reindex (and/or
> -recovery=2).
>
> not sure whether that's expected.
>
> my offer of trying to work on a "better" blob-enumeration still stand.
> but in the meantime, I'll try to setup the following scaffolding:
> - pk list -type=file > list.txt (on localhost)
> - for sha224 in list.txt; do pk-get -content $sha224 > filename
> - for filename in `ls -1`; do pk-upload remote-storage $filename
> (where pk-upload is just pk-put with remote-storage as the end-point and
> the following options: "file --filenodes --exiftime")

ok, perhaps unsurprisingly, this didn't pan out: even with a small
subsample (~30 images), I get the same error when restarting the
"remote-storage" server:

Starting perkeepd version master, 2022-07-10-c501f90cd0; Go go1.18.3 
(linux/arm64)
Starting to listen on http://localhost:3179
blobpacked: checking integrity of packed blobs against index...
blobpacked: 0 large blobs found in index, 29 missing from index
  sample missing large blob: sha224-098b...xxx
  sample missing large blob: sha224-1185...xxx
  ...

but, then, I tried something else:
- on "localhost", keep the same upload scaffolding as described above
- on "remote-storage", use the "default" server-config.json (ie: with
  sqlite instead of postgres)
and the "sample missing large blob: ...." errors disappeared.

what am I doing wrong?
looking at the output of \l in psql, I do see the "perkeep",
"pk_<id-lowercase>_blobpacked" and "pk_<id-lowercase>_syncto_index"
databases.

I have attached the relevant parts of the psql-based server (redacted) config.

please let me know what's amiss.

cheers,
-s

-- 
You received this message because you are subscribed to the Google Groups 
"Perkeep" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To view this discussion on the web visit 
https://groups.google.com/d/msgid/perkeep/CLI9KTXZB9NK.1HE83SUZ6ZV60%40zoidberg.

Attachment: pk-psql-config.json
Description: application/json

Reply via email to