Timeline for LastModifiedDate vs. SystemModstamp
Current License: CC BY-SA 3.0
10 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Dec 9, 2014 at 16:08 | vote | accept | Jim Horn | ||
| Dec 9, 2014 at 16:08 | comment | added | Jim Horn | Oaky. This would explain why the package I inherited detects changes based on Id-SystemModdate and not Id-LastModifiedDate. I'll have to investgate what if any 'Time based workflows, future methods, and workflow rule' our admins are doing that may change values in any manner other than user data entry. Thanks. | |
| Dec 9, 2014 at 14:58 | comment | added | Eric | @JimHorn - YES, Time based workflows, future methods, and workflow rule (small change in seconds I am assuming), trigger, etc.....Basically user clicks save, LastModifide updates, then triggers and everything else fires, SystemModStamp change (which should be different in seconds from last modified date, future, batch, etc may be widely different | |
| Dec 9, 2014 at 13:44 | comment | added | Jim Horn | Ok, we're getting very close. How about the inverse: Would there ever be a case where a value changed in any column (automated or otherwise), which would result in a change in SystemModstamp, but not a change in LastModifiedDate? This answer would nail it down for me. | |
| Dec 8, 2014 at 23:11 | comment | added | Eric | @JimHorn - No you cannot. I can click edit, then save, both values will change but no fields were actually changed | |
| Dec 8, 2014 at 14:30 | comment | added | Jim Horn | My assumption in doing the above, which is the heart of my question, is that if SystemModstamp changed can I assume that at least one value changed, and conversely if SystemModstamp did not change can I assume that the entire row has not changed? | |
| Dec 8, 2014 at 14:29 | comment | added | Jim Horn | What I'm trying to do is capture every change on my end. Since the SF rows are unique (that's a Type 1 SCD) on my end I'm trying to interpret if the row changed by comparing id and SystemModstamp to what I currently have in SQL Server. If SystemModstamp changed for a given id then there was a change, which I'm storing in SQL as a Type 2 SCD with a start_dt (SystemModstamp,) and end_dt (one second before next row's start_dt, or NULL if current row). | |
| Dec 5, 2014 at 22:04 | comment | added | Eric | @JimHorn - If you are using the ID with something else, how could you have duplicates as the ID is unique???? You could potentially have rows that do not match but the SF ID exists due to data skew. It all depends on your process flow | |
| Dec 5, 2014 at 21:07 | comment | added | Jim Horn | Ok. So then the question becomes .. 'Will the SystemModstamp value ever change if there is not a change in any other columns? The reason I ask is because in my current design I'm using the combination id and SystemModstamp as the primary key, assuming that if any data changes it will result in a new SystemModstamp value. However, I need to know if I'm running the risk of duplicate rows where two rows have all the same values except for the SystemModstamp. | |
| Dec 5, 2014 at 0:27 | history | answered | Eric | CC BY-SA 3.0 |