alibaba/DataX · error · IllegalArgumentException

to or from missed in edge

Error message

to or from missed in edge

What it means

In gdbwriter's DefaultGdbMapper, for EDGE import type each record must produce non-null 'to' and 'from' vertex ids (resolved via forStrColumn from the configured column rules). If either resolves to null, the mapper throws IllegalArgumentException after logging the invalid record — an edge cannot exist without both endpoint ids.

Source

Thrown at gdbwriter/src/main/java/com/alibaba/datax/plugin/writer/gdbwriter/mapping/DefaultGdbMapper.java:96

                    if (key == null || name == null) {
                        continue;
                    }
                    addToProperties(e, key, name, type, card);
                }
            });
        }

        final BiConsumer<Record, GdbElement> ret = (r, e) -> {
            final String label = forStrColumn(numPattern, rule.getLabel()).apply(r);
            String id = forStrColumn(numPattern, rule.getId()).apply(r);

            if (rule.getImportType() == Key.ImportType.EDGE) {
                final String to = forStrColumn(numPattern, rule.getTo()).apply(r);
                final String from = forStrColumn(numPattern, rule.getFrom()).apply(r);
                if (to == null || from == null) {
                    log.error("invalid record to: {} , from: {}", to, from);
                    throw new IllegalArgumentException("to or from missed in edge");
                }
                ((GdbEdge)e).setTo(to);
                ((GdbEdge)e).setFrom(from);

                // generate UUID for edge
                if (id == null) {
                    id = UUID.randomUUID().toString();
                }
            }

            if (id == null || label == null) {
                log.error("invalid record id: {} , label: {}", id, label);
                throw new IllegalArgumentException("id or label missed");
            }

            e.setId(id);
            e.setLabel(label);

View on GitHub (pinned to 80ec23d5c5)

Solutions

1. Check the logged 'invalid record to: {}, from: {}' line to identify which side is null and for which record
2. Fix null endpoint ids at the source or filter such rows (reader where-filter) before the writer
3. Verify the writer's 'column' mapping: to/from entries must reference existing columns carrying vertex ids
4. If null to/from is legitimately possible, decide policy (skip row) — the plugin has no skip option, so pre-filter

Example fix

// before: reader emits rows where fk can be null
// after: filter at read time
"where": "fk_to is not null and fk_from is null-free-equivalent"

// filter rows lacking endpoint ids before the writer sees them (reader-level)
"where": "fk_to IS NOT NULL AND fk_from IS NOT NULL"

catch (IllegalArgumentException e) {
    if ("to or from missed in edge".equals(e.getMessage())) {
        // log already printed the offending to/from; re-run with a where-filter to skip such rows
    }
    throw e;
}

When it happens

Trigger: gdbwriter job with importType=EDGE where the record's to/from column values are null, or the configured column index/name mapping for to/from does not exist so the rule resolves null. Also a '$to'/'$from' placeholder typo in the mapping JSON.

Common situations: Source rows with NULL foreign keys; writer column list order not matching the reader's column order so to/from point at wrong (possibly null) columns; mapping JSON referencing renamed columns after upstream schema change.

Related errors

• id or label missed
• Failed to cast %s to Long
• Failed to cast %s to Double
• Failed to cast %s to Boolean
• File not found: ${name}

AI-assisted analysis of alibaba/DataX@80ec23d5c5 (2026-08-14). Data as JSON: /api/errors/2b3193fa9b95eabf. Report an issue: GitHub.