High Note regarding fix
This issue has been fixed in Crowd 3.4.3 as well as 3.5.0. However those Crowd versions will only set the correct column types on newly created columns. If upgrading from an affected version (3.3.0-3.4.0) it's important to check for varchar, clob and text columns in the database and manually change them to the appropriate types.
Issue Summary
Prior to 3.3, customers could use our custom SQLServerIntlDialect and tables would be created as nvarchar, instead of MS SQL default varchar.
Starting with Crowd 3.,3 db schema creation process is using liquibase, as a result, the dialect is not respected anymore and tables being created as varchar or converted to varchar during an upgrade.
Environment
Crowd 3.3
Ms SQL
Expected results
When using SQLServerIntlDialect dialect, tables are created with nvarchar columns allowing them to store Unicode characters.
Actual Results
Tables are created with MSSQL default varchar
During an upgrade from earlier versions, if columns were nvarchar, the upgrade will convert them to varchar replacing Unicode characters already stored in the tables in the process.
Workaround
There is no workaround to prevent this, but columns can be manually changed to nvarchar after the upgrade/installation.
![[JIRA Server (Bulldog)]](/images/icons/generic_link_16.png "[JIRA Server (Bulldog)]"){kind=icon}
![[Atlassian Documentation] Page](https://confluence.atlassian.com/images/icons/favicon.png "[Atlassian Documentation] Page"){kind=icon}
[CWD-5358] During db schema creation on 3.3 all tables will be created in varchar
| Remote Link | Original: This issue links to "Page (Confluence)" [ 431295 ] |
| Remote Link | Original: This issue links to "Page (Confluence)" [ 630369 ] |
| Remote Link | New: This issue links to "Page (Confluence)" [ 630369 ] |
| Remote Link | New: This issue links to "Page (Confluence)" [ 630258 ] |
| Epic Link | Original: CWD-4704 [ 600140 ] |
| Workflow | Original: Simplified Crowd Development Workflow v2 - restricted [ 3097688 ] | New: JAC Bug Workflow v3 [ 3366046 ] |
| Status | Original: Resolved [ 5 ] | New: Closed [ 6 ] |
| Remote Link | New: This issue links to "Page (Atlassian Documentation)" [ 432508 ] |
| Description |
Original:
h3. Note regarding fix
This issue has been fixed in Crowd 3.4.3 as well as 3.5.0. However those Crowd versions will only set the correct column types on newly created columns. If upgrading from an affected version (3.3.0-3.4.2) it's important to check for varchar, clob and text columns in the database and *manually change them to the appropriate types.* h3. Issue Summary Prior to 3.3, customers could use our custom *SQLServerIntlDialect* and tables would be created as *nvarchar*, instead of MS SQL default *varchar*. Starting with Crowd 3.,3 db schema creation process is using liquibase, as a result, the dialect is not respected anymore and tables being created as varchar or converted to varchar during an upgrade. h3. Environment Crowd 3.3 Ms SQL h3. Expected results When using *SQLServerIntlDialect* dialect, tables are created with nvarchar columns allowing them to store Unicode characters. h3. Actual Results Tables are created with MSSQL default varchar During an upgrade from earlier versions, if columns were nvarchar, the upgrade will convert them to varchar replacing Unicode characters already stored in the tables in the process. h3. Workaround There is no workaround to prevent this, but columns can be manually changed to nvarchar after the upgrade/installation. |
New:
h3. Note regarding fix
This issue has been fixed in Crowd 3.4.3 as well as 3.5.0. However those Crowd versions will only set the correct column types on newly created columns. If upgrading from an affected version (3.3.0-3.4.0) it's important to check for varchar, clob and text columns in the database and *manually change them to the appropriate types.* h3. Issue Summary Prior to 3.3, customers could use our custom *SQLServerIntlDialect* and tables would be created as *nvarchar*, instead of MS SQL default *varchar*. Starting with Crowd 3.,3 db schema creation process is using liquibase, as a result, the dialect is not respected anymore and tables being created as varchar or converted to varchar during an upgrade. h3. Environment Crowd 3.3 Ms SQL h3. Expected results When using *SQLServerIntlDialect* dialect, tables are created with nvarchar columns allowing them to store Unicode characters. h3. Actual Results Tables are created with MSSQL default varchar During an upgrade from earlier versions, if columns were nvarchar, the upgrade will convert them to varchar replacing Unicode characters already stored in the tables in the process. h3. Workaround There is no workaround to prevent this, but columns can be manually changed to nvarchar after the upgrade/installation. |
| Description |
Original:
h3. Issue Summary
Prior to 3.3, customers could use our custom *SQLServerIntlDialect* and tables would be created as *nvarchar*, instead of MS SQL default *varchar*. Starting with Crowd 3.,3 db schema creation process is using liquibase, as a result, the dialect is not respected anymore and tables being created as varchar or converted to varchar during an upgrade. h3. Environment Crowd 3.3 Ms SQL h3. Expected results When using *SQLServerIntlDialect* dialect, tables are created with nvarchar columns allowing them to store Unicode characters. h3. Actual Results Tables are created with MSSQL default varchar During an upgrade from earlier versions, if columns were nvarchar, the upgrade will convert them to varchar replacing Unicode characters already stored in the tables in the process. h3. Workaround There is no workaround to prevent this, but columns can be manually changed to nvarchar after the upgrade/installation. |
New:
h3. Note regarding fix
This issue has been fixed in Crowd 3.4.3 as well as 3.5.0. However those Crowd versions will only set the correct column types on newly created columns. If upgrading from an affected version (3.3.0-3.4.2) it's important to check for varchar, clob and text columns in the database and *manually change them to the appropriate types.* h3. Issue Summary Prior to 3.3, customers could use our custom *SQLServerIntlDialect* and tables would be created as *nvarchar*, instead of MS SQL default *varchar*. Starting with Crowd 3.,3 db schema creation process is using liquibase, as a result, the dialect is not respected anymore and tables being created as varchar or converted to varchar during an upgrade. h3. Environment Crowd 3.3 Ms SQL h3. Expected results When using *SQLServerIntlDialect* dialect, tables are created with nvarchar columns allowing them to store Unicode characters. h3. Actual Results Tables are created with MSSQL default varchar During an upgrade from earlier versions, if columns were nvarchar, the upgrade will convert them to varchar replacing Unicode characters already stored in the tables in the process. h3. Workaround There is no workaround to prevent this, but columns can be manually changed to nvarchar after the upgrade/installation. |
| Remote Link | New: This issue links to "Page (Confluence)" [ 431295 ] |