[Docs audit] Concept pages contain factual errors (MariaDB, incremental definition, implementation details, glossary, FAQ, peripheral pages)
I maintainer di solito rispondono entro 1 giorno
Nessuno ha ancora preso questa issue.
Valutazione
- Difficoltà
- 4/5
- Tempo stimato
- 1-2 giorni
- Idoneità per principianti
- 45/100
- Tipo di issue
- Documentazione
- Chiarezza
- Abbastanza chiara
- Stato di attività
- Attiva
- Stack tecnologico
- mariadb
- Ambito
- databases, documentation
Direzione di ricerca
Start with docs/backup-overview.md line 3 on the 8.4 branch, then check the same claims in differences-logical-backup, features, about-xtrabackup, xtrabackup-implementation-details, glossary and faq, and compare the incremental-backup wording with backup-strategy.md, which the issue says is already correct. Peripheral fixes are in verify-backup.md and the update-curl-utility and toolkit-version-check pages. Done means each listed wrong statement is corrected on the affected branches (8.0, 8.4, 9.7) and the replacement texts F20, F21 and F24 are applied, though those IDs are not in this payload.
Scritto dal modello di indicizzazione a partire dal testo della issue.
Descrizione
Found during a hands-on audit of the Percona XtraBackup docs (2026-10-12): every page was executed as written on Ubuntu 24.04, Ubuntu 26.04, Debian 13, Rocky 9 and Rocky 10 (SELinux enforcing) with PXB 8.0.35-36, 8.4.0-7 and 9.7.1-rc1 against Percona Server 8.0.46/8.4.11/9.7.2 and Oracle MySQL 8.0.46/8.4.11/9.7.2, with data and a concurrent write load; the rendered sites were also crawled and link-checked. Source references are <branch>:docs/<file>.md:<line> (branches 8.0 @6145834, 8.4 @5d943a6, 9.7 @5aff2c7). Finding IDs (D1-D70) are shared across the issues of this audit:
Where each finding is filed
| Issue | Findings |
|---|---|
| #512 | D30, D31, D40, D41 |
| #513 | D1, D53, D64 |
| #514 | D2, D19 |
| #515 | D44, D45 |
| #516 | D5 |
| #517 | D32, D47, D61, D28 |
| #518 | D17, D12, D18, D38 |
| #519 | D6, D35, D27, D24 |
| #520 | D14, D26, D29, D33, D23 |
| #521 | D3, D7, D4, D8 |
| #522 | D9, D10, D11, D34, D39, D62 |
| #523 | D20, D21, D22, D36, D15, D42, D43, D49, D37 |
| #524 | D13, D16 |
| #525 | D54, D55, D56, D57, D58, D59, D63 |
| #526 | D60, D51, D52 |
| #527 | D46, D50 |
| #528 | D65, D66, D67 |
| #529 | D68, D69, D70, D48, D25 |
Findings in this issue: D60 (Medium), D51 (Low), D52 (Low)
D60. Concept pages contain factual errors: MariaDB, incremental and partial definitions, "no load", implementation details, glossary, FAQ
- Severity: Medium
- Series: 8.4, 9.7 (backup-overview also in 8.0)
- Pages:
- 8.4/backup-overview (
:3-15) - differences-logical-backup (
:8,19,27) - features (
:15) - about-xtrabackup (
:7-19) - xtrabackup-implementation-details (
:8-13,39-44) - glossary
- faq (
:6-68)
- 8.4/backup-overview (
- What happened:
- Wrong scope and mechanics:
- PXB does not support MariaDB.
- It copies only the current binlog fragment, not "the binary log files".
- Wrong definitions:
- An incremental backup copies pages changed since the previous backup, full or incremental. A verified chain shows inc2's from_lsn equals inc1's to_lsn, and backup-strategy.md already says this correctly.
- A partial backup is a set of selected tables or databases, not "files changed in a folder".
- Logical dumps can include users.
- Overstated claims:
- "No load": strace shows 10 MiB reads.
- "Non-blocking": non-InnoDB writes are blocked on PS.
- "Verify backup automatically": no such feature exists.
- Implementation details are wrong:
- PXB needs read access only; it ran as a group-read OS user.
- Reads are
--read-buffer-size, 10 MiB by default and configurable, not "1 MB non-configurable". - Redo is read in 64 KiB chunks.
innodb_log_block_sizedoes not exist.
- Glossary errors:
.frm,.par,.TRG,.TRNand.optdo not exist in 8.x.- The data dictionary lives in
mysql.ibd. - ibdata1 is 12 MB, not 10 MB.
- The variable is
innodb_log_group_home_dir, notinnodb_log_group_home.
- FAQ errors:
- The "1048576 hard-coded" claim is wrong; the limit is
fs.nr_open, 1073741816 on Rocky 10. - The ZRM link returns 403.
- The optimized-DDL error text differs from 8.4's.
- The "1048576 hard-coded" claim is wrong; the limit is
- Wrong scope and mechanics:
- Fix: apply the replacement texts in F20, F21, F24 and . For example, backup-overview line 3 becomes: "...physical backups of MySQL and Percona Server for MySQL (MariaDB is not supported; use mariabackup). It copies InnoDB data files and the redo log generated during the copy, plus non-InnoDB files and the current binary log fragment."
D51. Peripheral pages are outdated: verify-backup, update-curl-utility, toolkit-version-check
- Severity: Low
- Series: 8.4 (verify-backup in all series)
- What happened:
- verify-backup (
8.4:verify-backup.md:4,11-13,28,33,80):- percona.com/doc links soft-404 to the docs home page;
GRANT ... IDENTIFIED BYfails on 8.x;- the replication guide is not linked.
- update-curl-utility:
- it checks the curl CLI instead of the library xbcloud loads;
apt install --reinstallcannot downgrade;- the curl 7.88.1 sample with "Metalink" is fabricated (7.88.1 was released on 2023-02-20 and has no Metalink);
- it gives Debian 12 instructions only;
- the page title and nav label disagree.
- toolkit-version-check:
- PXB 8.4 performs no version check (strace shows only connects to the local socket);
--no-version-checkis a deprecated no-op.- PXB 8.0 does run the check, per .
- verify-backup (
- Fix:
- verify-backup: use https://docs.percona.com/percona-toolkit/pt-table-checksum.html and
CREATE USER. - update-curl-utility: check with
ldd $(which xbcloud) | grep libcurlanddpkg -l 'libcurl4*'/rpm -q libcurl, give an explicit-version downgrade, and add a per-distribution table. - toolkit-version-check: "PXB 8.4 does not perform a version check".
- verify-backup: use https://docs.percona.com/percona-toolkit/pt-table-checksum.html and
D52. Mixed slave/master and replica/source terminology
- Severity: Low
- Series: 8.0, 8.4, 9.7
- What happened: 8.4 and 9.7 use
--slave-info,--safe-slave-backupandxtrabackup_slave_infonext to "replica" prose. 8.0 uses CHANGE MASTER, START SLAVE and SHOW SLAVE STATUS. - Fix: add a Terminology note mapping slave to replica and master to source, stating which option names each series accepts.
Found in a hands-on docs audit run with Claude Code.
- Lingua principale
- HTML
- Stelle
- 20
- Fork
- 40
- Merge medio
- 17h 14m
- PR unite (30g)
- 11
Preparare l'ambiente
- Nessun Dockerfile né file Docker Compose
- Nessun modello di pull request
- Leggi la guida per i contributori
Come iniziare
- Leggi tutta la issue e poi la guida ai contributi del progetto.
- Commenta sulla issue per dire che te ne occupi tu — evita che due persone facciano lo stesso lavoro.
- Fai un fork del repository e lavora su un branch.
- Apri una pull request che faccia riferimento al numero della issue.
Altre issue di percona/pxb-docs
-
[Docs audit] Rendering leftovers, broken links/anchors, stale sample output, compile page, typos (bulk)Forse già presa @alina-derkach-oaza l’ha presa 11 giorni fa. Aperta
percona/pxb-docs#529 · 1 assegnatario ·
I maintainer di solito rispondono entro 1 giorno
-
Difficoltà 4/5 3-5 giorni Idoneità per principianti 25/100
I maintainer di solito rispondono entro 1 giorno
-
Difficoltà 3/5 Mezza giornata Idoneità per principianti 55/100
I maintainer di solito rispondono entro 1 giorno
-
Difficoltà 4/5 1-2 giorni Idoneità per principianti 35/100
I maintainer di solito rispondono entro 1 giorno
-
Difficoltà 4/5 3-5 giorni Idoneità per principianti 30/100
I maintainer di solito rispondono entro 1 giorno
Tutte le issue di percona/pxb-docs
Issue simili
-
[BUG] dcg.waggr permutation p-values are folded twice, so near-zero scores get the smallest p-valuesAperta
Difficoltà 2/5 1-3 ore Idoneità per principianti 75/100
scverse/rapids-singlecell#854 ·
I maintainer di solito rispondono entro 1 giorno
-
SQL detection rejects valid statements with extra whitespace around tokens or after the semicolonForse già presa @PerryLink l’ha presa oggi. Aperta
Difficoltà 2/5 1-3 ore Idoneità per principianti 72/100
I maintainer di solito rispondono entro 1 giorno
-
area-testing p1-recommended
Difficoltà 2/5 1-3 ore Idoneità per principianti 68/100
I maintainer di solito rispondono entro 1 giorno
-
Unblock: Kanal11.seApertablocklist:remove check:passed
Difficoltà 2/5 1-3 ore Idoneità per principianti 63/100
-
checkpoint/verify/rebuild_indexes on a thread holding a WriteTransaction deadlock instead of INT-001Apertabug known-issue
Difficoltà 2/5 1-3 ore Idoneità per principianti 74/100
project-minigraf/minigraf#734 · 1 commento ·
I maintainer di solito rispondono entro 1 giorno