Summary: | ASTERISK-24283: [patch]Microseconds precision in the eventtime column in the cel_odbc module | ||
Reporter: | Etienne Lessard (hexanol) | Labels: | |
Date Opened: | 2014-08-28 14:37:32 | Date Closed: | 2014-11-15 10:54:45.000-0600 |
Priority: | Minor | Regression? | |
Status: | Closed/Complete | Components: | CEL/cel_odbc |
Versions: | 11.11.0 | Frequency of Occurrence | |
Related Issues: | |||
Environment: | Attachments: | ( 0) cel_odbc_time_precision.patch | |
Description: | In the cel_odbc.conf.sample file, it's written:
{noformat} ... ; FORWARD = 25 ; eventtime (timeval, includes microseconds) ; userdeftype (set only if eventtype == USER_DEFINED) ... {noformat} That said, if you use cel_odbc to store CEL in a table where the eventtime column is of timestamp type, you'll only get second precision, i.e. you won't have sub-second precision. The attached patch fix that. Tested on asterisk 11.11.0 with a postgresql 9.1 DBMS with a CEL table with the eventtime column of type timestamp. Also, if you were using an eventtime column of type varchar, I believe the patch will also increase the precision of the data stored; you'll go from millisecond precision to microsecond precision. That said, personally, it would be fine to have millisecond precision; microsecond precision is not really important. | ||
Comments: | By: Rusty Newton (rnewton) 2014-09-04 20:29:10.542-0500 [~hexanol] Thanks! Can you put this on Reviewboard as well to help push it along through the [Code Review|https://wiki.asterisk.org/wiki/display/AST/Code+Review] process? By: Etienne Lessard (hexanol) 2014-09-05 14:55:02.622-0500 Yes, I just did: https://reviewboard.asterisk.org/r/3980/ |