Add WACZ filename, depth, favIconUrl, isSeed to pages

[tw4l]

Fixes #2348

Adds filename to pages, pointed to the WACZ file those files come from.

Also adds idempotent migration to backfill this information for existing pages.

Update: Now also adding depth, isSeed, and favIconUrl to pages as well, both for crawls moving forward and in the backfill migration.

Manual testing

1. Spin up a local environment off of latest main
2. Create some crawls, verify pages are added to database
3. Re-deploy from this branch
4. Verify migration ran successfully in backend logs, didn't raise errors
5. Verify filename and other fields have been added for pages in older crawls
6. Create a new crawl, verify filename is present and matches values in backfilled pages (namely, that it's just the WACZ filename, with no oid directory prefix)

[tw4l]

Converted to draft to add some additional features:

• Adding depth, faviconUrl, and seed (potentially? duplicates depth a bit) to pages and to backfill migration
• Looking into ensuring long-running migrations finish without issue, perhaps in init_container or by modifying migration to spawn background jobs

[ikreymer]

Converted to draft to add some additional features:

• Adding depth, faviconUrl, and seed (potentially? duplicates depth a bit) to pages and to backfill migration
• Looking into ensuring long-running migrations finish without issue, perhaps in init_container or by modifying migration to spawn background jobs

Added depth so far. Should we add faviconUrl? Could potentially use it later.
Didn't add seed as that's just depth == 0, main reason would be for an index that we can filter on isSeed true/false, which we might want to do for querying. Though, could still do that with an index on depth.

Re: init_container, I think we can just bump the startupProbe max time for now as the easiest option.

[tw4l]

Added depth so far. Should we add faviconUrl? Could potentially use it later. Didn't add seed as that's just depth == 0, main reason would be for an index that we can filter on isSeed true/false, which we might want to do for querying. Though, could still do that with an index on depth.

It's such a small amount of additional information, I think it's probably worth the overhead to add all of these fields into the db if there's even a chance they'll be useful to have later (certainly I can see the favicon url being used before long!).

[tw4l]

@ikreymer this should be ready to go for re-review/testing on dev now, there is a test failure but it's unrelated (QA issue, maybe due to a crawler or wabac change)