Gluesync Knowledge Base: errors and fixes
Search this page for the code or the first words of the message you see, in a notification, in the Control Plane or in the logs. Each row says what the error means and links to the section of a guide that explains the fix step by step.
If your error is not here, follow Read Gluesync logs and diagnose errors, then Create a support ticket with the logs it asks for.
IBM i (AS400) sources
| Error or symptom | What it means | Fix |
|---|---|---|
|
The Gluesync user has |
Fix SQL0443 on a journal, authority checklist, authorities of the Gluesync user |
|
The user lacks |
|
|
The user lacks |
|
|
The Gluesync user lacks |
|
Missing Journal Table Definition, |
The table is not journaled, or is journaled with after images only |
|
Journal checkpoint outdated, |
A journal receiver that Gluesync still needed was deleted |
|
|
Earlier builds of the Validator selected the relative record number by name, which Db2 for i refuses |
|
|
The technical field is off for the connection, or the object is a logical file |
|
|
The IBM i query time limit stopped a Validator query |
|
A target column is too short for text from the IBM i |
EBCDIC text takes more bytes in UTF-8 on the target |
|
A snapshot is slow, or stops with |
A network device between Gluesync and the IBM i closes connections that stay idle during long reads |
|
|
Another job holds a lock on the file, for example during a reorganize. Gluesync retries |
|
One entity is hours late, or a pipeline with hundreds of entities drifts after a resync |
The entity polling interval is too high, or the shared journal reader is behind |
Oracle
| Error or symptom | What it means | Fix |
|---|---|---|
|
The value is longer than the target column. Often a column sized in bytes receives multibyte characters |
|
|
The source ran out of |
|
|
The Gluesync user is missing a grant for LogMiner |
|
|
The redo or archive log that holds the saved position is no longer available |
|
|
Connection, listener or database state errors |
SQL Server
| Error or symptom | What it means | Fix |
|---|---|---|
|
A value is longer than the target column |
|
The same truncation error, although the target column is wide enough |
In some releases the staging table of a large batch declared |
|
Other errors writing to a SQL Server target |
Permissions, collation and schema of the target |
MySQL and MariaDB
| Error or symptom | What it means | Fix |
|---|---|---|
The binlog format is not |
CDC needs row-based binlogs that are kept long enough |
|
The snapshot works but CDC changes never arrive |
CDC reads only the database set as Database name of the connection |
Validator
| Error or symptom | What it means | Fix |
|---|---|---|
|
The source returns rows in an order the comparison cannot use |
|
|
A comparison query or the whole run took longer than its limit |
|
A scheduled Validate job shows a red Last run |
One entity of the job failed and the next entities did not run |
|
|
Too many comparisons at the same time |
|
What a counter or a status of a run means |
||
The Validate page does not load |
The browser keeps an old version of the page |
|
|
The table has no usable key for the comparison |
|
A comparison of a big table times out, or a connector is not supported |
||
Thousands of Target only rows and |
A |
Snapshots and row counts
| Error or symptom | What it means | Fix |
|---|---|---|
The target has fewer rows than the source |
A bulk snapshot has not confirmed its first chunk yet, or the source holds the same key more than once |
|
|
In CDC with On duplicate key set to Skip, rows of one transaction that changed their keys met on the target, and Skip discarded them |
|
A snapshot is slow |
Page size, partitioning, bulk load and the source query plan |
Pipelines, alerts and upgrades
| Error or symptom | What it means | Fix |
|---|---|---|
A pipeline stays In maintenance |
A maintenance schedule did not resume it, or it was put in maintenance by hand |
|
Entity Configuration Changed: the entity "has been stopped", "a new resync is required" |
The keys or columns of the entity changed, so Gluesync stopped it until a resync |
|
Changes arrive, but late |
A very large transaction delays the changes behind it. Changes that wait longer than Source change retention in hours are dropped |
|
Informix |
The logical logs were replaced after the saved position |
|
High CPU on the Gluesync host or on the source |
Polling intervals, journal readers, connection pools, heap or host size |
|
Automator: |
The bulk run started before Core Hub reached the target database |
|
Automator: HTTP 500 on Import All, field functions or passwords missing after a restore, |
||
Email alerts do not arrive, or arrive too often |
SMTP settings, recipients or the digest settings |
|
Something behaves differently after an upgrade from Gluesync 2.2 to 2.3 |
SQL Anywhere
| Error or symptom | What it means | Fix |
|---|---|---|
|
The client folder is incomplete, not mounted or not the one the agent looks for |
|
|
A Core Hub process loads one client folder |
|
|
Core Hub runs on another architecture or operating system |
|
SQL Anywhere transaction log cannot be read, |
The user lacks |
|
The notification says the log was renamed or truncated past the position of the agent |
A log rename ran with commits that the agent had not read |
|
SQL Anywhere error |
A row with that key is already on the target |
Greenplum, Supabase and Elasticsearch targets
| Error or symptom | What it means | Fix |
|---|---|---|
|
The user cannot read the catalog view of external tables |
|
|
The user lacks the role attribute for readable |
|
|
gpfdist advertised host does not resolve from a segment host |
|
|
On Greenplum 6, a primary key on an append-optimized table that you create yourself |
|
|
The extension is not installed in the target database |
|
|
Wrong password or user name, or the project reference is missing on the session pooler |
|
|
The host cannot be reached from Core Hub. The direct connection is IPv6 only without the IPv4 add-on |
|
|
Trust server certificate is off and no trust store is set |
|
|
The role does not own the table and has no privilege on it |
|
|
The host name does not resolve from Core Hub |
|
|
Enable TLS is on, but the cluster answers in plain HTTP |
|
|
The user name or the password is wrong |
|
|
The role has no |
|
|
The role has no |
Automator
| Error or symptom | What it means | Fix |
|---|---|---|
|
Enable TLS is selected, and Core Hub answers on plain HTTP |
|
|
The target schema is missing, or its name has a typing error |
|
|
The database reached |
|
|
The archive was exported without Include secrets |
|
|
One entity could not be created |
Installer, Conductor and platform codes
| Error or symptom | What it means | Fix |
|---|---|---|
|
Codes of the pre-flight checks, the installers and Gluesync Control |
|
|
Diagnosis codes returned by Gluesync Conductor |
|
A CVE or a scanner finding about Gluesync |