asfsdafsad

Please take a look at my comment on JSWSERVER-15917.

This is not resolved, due to wrong data type is being declared for Quartz tables in in fieldtype-mssql.xml under \WEB-INF\classes\entitydefs

Please reopen this bug,as this isssue is happening in 7.2.7

PS - do not do the workaround, you will no longer be able to paste non ascii characters into your comment fields.

Hi,

We have put significant amount of work to make ranking faster when working on JSW-13163.
It makes the operation much faster also on Microsoft SQL Server.

Therefore I'm closing the issue as fixed. In case of any performance problems, please leave a comment here or ask our support for help and investigation.

Thanks,
Jacek Jaroczynski
JIRA Bugmaster
[Atlassian]

When we will have a fix for this. How long we have to run

DROP INDEX [index_ao_60db71_lexorank_rank] ON [dbo].[AO_60DB71_LEXORANK]
GO
 
alter table [AO_60DB71_LEXORANK] alter column [rank] nvarchar(255)
GO
 
CREATE NONCLUSTERED INDEX [index_ao_60db71_lexorank_rank] ON [dbo].[AO_60DB71_LEXORANK]
(
      [RANK] ASC
)WITH (PAD_INDEX = OFF, STATISTICS_NORECOMPUTE = OFF, SORT_IN_TEMPDB = OFF, DROP_EXISTING = OFF, ONLINE = OFF, ALLOW_ROW_LOCKS = ON, ALLOW_PAGE_LOCKS = ON) ON [PRIMARY]
GO 

Soon after restarting JIRA.
Support took 3 months to flag this as a bug and its about to be 5 months passed, since then.

vkharisma

You missed the statement under the work around.
if we set "sendStringParametersAsUnicode=false", then JIRA failing to work with non English languages. If I enter Russian text in comments, it saves as "???".

At this time, we are using JIRA Agile 6.7.11, with MSSQL 2008 R2, 64 Bit Database server.
And still we are having the issue. Any resolution or plan to fix this one yet.

I would like to update a test report on this context.

Observation:

1. This issue is conversion of String to Unicode and Compare, and the performance overhead on JDBC level.
2. jTDS JDBC driver do not handle that situation well. To handle that, if we set "sendStringParametersAsUnicode=false", then JIRA failing to work with non English languages. If I enter Russian text in comments, it saves as "???".
3. As document says, JDBC 4 Spec, handles that well.

So Test Situations:

1. I converted Rank column of [AO_60DB71_LEXORANK], to nvarchar(255) from varchar(255) and tested with default jTDS driver and Microsoft JDBC driver 4.2. and found that the rank performance issue is gone, when i run that with JDBC 4.2 driver.

What I did.

1. Copied "sqljdbc42.jar" file to JIRA install "\JIRA\lib", where "jtds-1.3.1.jar" do exist.
2. changed dbconfig.xml file as

    <url>jdbc:sqlserver://localhost:1433;databaseName=jira6test;user=xxxx;password=yyyy;</url>
   		<driver-class>com.microsoft.sqlserver.jdbc.SQLServerDriver</driver-class> 

3. Started JIRA.
4. Run this SQL against the database jira6test

    
   DROP INDEX [index_ao_60db71_lexorank_rank] ON [dbo].[AO_60DB71_LEXORANK]
   GO
   
   alter table [AO_60DB71_LEXORANK] alter column [rank] nvarchar(255)
   GO
   
   CREATE NONCLUSTERED INDEX [index_ao_60db71_lexorank_rank] ON [dbo].[AO_60DB71_LEXORANK]
   (
   	[RANK] ASC
   )WITH (PAD_INDEX = OFF, STATISTICS_NORECOMPUTE = OFF, SORT_IN_TEMPDB = OFF, DROP_EXISTING = OFF, ONLINE = OFF, ALLOW_ROW_LOCKS = ON, ALLOW_PAGE_LOCKS = ON) ON [PRIMARY]
   GO 

And the Tested, and found that the issue is gone.

Converting to nvarchar(255), do not give any benefit with jTDS driver.

Yes, we do have this bug. And we are expecting a fix.