Add support for datasource properties about jcliff HOT 9 CLOSED

These are not JDBC driver properties, they are datasource properties, and
setting them is already supported. Please let me know if this doesn't work.
Thanks.

On Mon, Jul 27, 2015 at 4:13 AM, mschoepke [email protected] wrote:

Please add support for the following properties of a JDBC driver:

• validation
  • check-valid-connection-sql
  • validate-on-match
  • background-validation
  • background-validation-millis
  • ...
    • timeout
  • set-tx-query-timeout
  • blocking-timeout-millis
  • idle-timeout-minutes
  • query-timeout
  • use-try-lock
  • allocation-retry
  • allocation-retry-wait-millis
  • ...

—
Reply to this email directly or view it on GitHub
#54.

from jcliff.

Yes, it works fine, thanks.

from jcliff.

I just found out that using datasource properties in jcliff breaks its Idempotence. jcliff will reload every time it runs.

Example: background-validation-millis should be set to 30000. If I read the log correctly, it finds it to be set to 30000L. Don't know where the L comes from. Of course, it will write the attribute again to be 30000.

Any clue?

from jcliff.

It is probably because of automatic type conversion. You set some value to
30000, it is internally a long value, so it shows up as 30000L (L means
"long"), jcliff compares the nodes and sees "30000"!="30000L", and applies
the rule again.

If jboss-cli says it is 30000L, then set it to 30000L in your configuration
file, and it should work once. I'll try to find a way to change the
comparison algorithm to ignore such syntax differences.

On Thu, Nov 5, 2015 at 6:41 AM, MarS919 [email protected] wrote:

I just found out that using datasource properties in jcliff breaks its
Idempotence. jcliff will reload every time it runs.

Example: background-validation-millis should be set to 30000. If I read
the log correctly, it finds it to be set to 30000_L_. Don't know where
the L comes from. Of course, it will write the attribute again to be 30000.

Any clue?

—
Reply to this email directly or view it on GitHub
#54 (comment).

from jcliff.

No, jboss-cli says it's 30000. Just jcliff reads it as 30000L.

Setting it to 30000L in my configuration results in

Exception in thread "main" java.io.IOException: Invalid character: L

from jcliff.

ok, i'll check

On Thu, Nov 12, 2015 at 7:31 AM, MarS919 [email protected] wrote:

No, jboss-cli says it's 30000. Just jcliff reads it as 30000L.

Setting it to 30000L in my configuration results in

Exception in thread "main" java.io.IOException: Invalid character: L

—
Reply to this email directly or view it on GitHub
#54 (comment).

from jcliff.

Please try with 2.10.13.

On Thu, Nov 12, 2015 at 7:32 AM, Burak Serdar [email protected] wrote:

ok, i'll check

On Thu, Nov 12, 2015 at 7:31 AM, MarS919 [email protected] wrote:

No, jboss-cli says it's 30000. Just jcliff reads it as 30000L.

Setting it to 30000L in my configuration results in

Exception in thread "main" java.io.IOException: Invalid character: L

—
Reply to this email directly or view it on GitHub
#54 (comment).

from jcliff.

Thanks for the new version, but the issue remains the same. Tell me when you need more information from me.

from jcliff.

Can you send me the full execution log when jcliff is run with this
configuration file causing the problem? Run jcliff with -v, and remove any
passwords in the output please.

Thanks.

On Mon, Nov 16, 2015 at 12:09 AM, MarS919 [email protected] wrote:

Thanks for the new version, but the issue remains the same. Tell me when
you need more information from me.

—
Reply to this email directly or view it on GitHub
#54 (comment).

from jcliff.