While working on Salesforce application a thought came into my mind about the data loss. Can anyone list the various scenarios in which one can loss the application data.
Not sure if I misunderstood, but these are some ways to prevent data loss:
- Physical damage to only place the data is stored i.e no backups have been taken so cannot be restored
- Human error - this could go on, but I remember someone dropping all the tables in a SQL database once by accident! Which leads me to my next point…
- Not restricting access to where the data is stored. Don’t let someone run wild on a live database
- Lack on monitoring on your hard drives or any component that could fail and cause data loss
I’m sure there are many more but this is just what I thought of off the top of my head.
You already have gotten a few good suggestions. Some additional things that I thought of.
To add to the first, any outage on any part of the transition for the data from the client to the database.
Another area is when data is updated, either through client updates or as side effects of other actions.
I would also like to add when software is updated, structure of the database is changed or other operational cases.
Finally a tricky one that is when you have a redundant system and several instances all keep portion of the state and has to do a handover. I.e. when your cluster is out of sync, if applicable.
In Salesforce, the data loss can happen in following situations:
- Salesforce supports very powerful integration tools, which can update a large number of records even in millions through automated processing. That means any problem in the integration could inadvertently update the wrong records. Without having the current backup, any error in integration could become impossible to recover.
- As we all know that SOQL is specifically designed for Salesforce and there are instant chances that SOQL mistakes can happen. Thus, the accidental use of wrong syntax can create a major problem. So, taking backup should be done prior to execution of any SQQL statements.
- At the time of data import, an incorrect mapping can lead to bad values in a large number of records.
So, it’s a good practice not only for the organizations which are using Salesforce application but also which are providing QA services to be aware of above situations so that any data loss can be avoided.
Hope this information is helpful for you.
When setting up your database, incorrectly identifying data fields can lead to lots of unknown issues (e.g., primary vs. foreign key identification and duplication of data that must be unique).