Quarterly database

The classic per-TLD layout of the previous product generation (gzip-compressed CSV, full / regular / simple formats), released quarterly, for scripts written against that layout.

Data availability

  • Update frequency: quarterly, independently of the monthly releases.
  • Folders: quarterly/legacy/whois_database/v%N%/ (generic TLDs) and quarterly/legacy/domain_list_quarterly/v%N%/ (country-code TLDs from the IANA root-zone list, including IDN ccTLDs such as xn--p1ai).
  • Current versions: gTLD v57 (1,130 TLDs, 671GiB), ccTLD v43 (248 TLDs, 139GiB).
  • Reference manual: Quarterly WHOIS Database Reference Manual
  • Download scripts: whois_database_download_support

Folder structure

whois_database/v57/                          # domain_list_quarterly/v43/ has the same shape
├── README.txt | README.html | README.pdf     # per-release manual: sizes, listings, coverage, top stats
├── docs/
│   ├── v57.tlds                              # all TLDs in the release, one comma-separated line
│   └── v57.csvs                              # same content
├── v43.csvs                                  # ccTLD tree only: the same list at the root
└── csv/
    ├── domains/<tld>/
    │   ├── domain_names_v57_<tld>.csv.gz             # one domain per line, no header
    │   ├── verified_domain_names_v57_<tld>.csv.gz    # identical to domain_names_
    │   ├── missing_domain_names_v57_<tld>.csv.gz     # empty (see below)
    │   ├── reserved_domain_names_v57_<tld>.csv.gz    # empty
    │   ├── thin_domain_names_v57_<tld>.csv.gz        # empty
    │   └── *.md5, *.sha256 for each
    ├── tlds/{full,regular,simple}/
    │   └── csvs.<tld>.<format>.tar.gz (+ .md5, .sha256)   # one tarball per TLD per format
    └── tlds_combined/
        └── <format>-v<N>.tar.gz.NNNN (+ .md5, .sha256)     # every TLD of the tree in one tar, split in 1GiB parts

Naming convention

File Description
Per-TLD tarball csv/tlds/%FORMAT%/csvs.%TLD%.%FORMAT%.tar.gz, %FORMAT% = full | regular | simple. Unpacks to %FORMAT%/%TLD%/%N%.csv: 100,000 data rows per file (1.csv, 2.csv, …), header in every file.
Combined parts csv/tlds_combined/%FORMAT%-v%N%.tar.gz.%NNNN%: slices of one tar stream with every TLD of the tree, not standalone archives. Concatenate before extracting: cat full-v57.tar.gz.* | tar xzf -. Counts in v57 / v43: gTLD full 271 / regular 36 / simple 29 parts; ccTLD 48 / 12 / 10.
Domain lists csv/domains/%TLD%/domain_names_v%N%_%TLD%.csv.gz, one domain per line, no header. The missing_, reserved_ and thin_ lists are empty, because this database lists only domains it holds a WHOIS record for; README section 4.1 shows 100% coverage for every TLD.
Checksums Every data file has .md5 and .sha256 sidecars, compatible with md5sum -c / sha256sum -c.

CSV structure

The formats have the same columns as the CSV variants of the Monthly database: full = 96 columns, regular = 94 columns (without RegistryData_rawText, WhoisRecord_rawText), simple = 43 columns (additionally without every *_rawText, the billing/technical/zone contacts and registrarIANAID). Column descriptions are on the Monthly database page.

Files are gzip-compressed: use unpigz rather than gunzip, which can be almost 2x slower on large files.