Hard Drive Crashed - Software Reinstalled - Validation req?


A validated system’s hard drive crashed and it is being replaced by a new one and the software will be reinstalled. How extensive should be the testing after the re-install and what all should be tested?


In brief. It really depends on what went wrong, what you are changing out, and what the differences are in what you are changing out.

An assessment should be done internally. In my experience, changes fit under 3 categories (1) LIke for like (2) Equivalent and (3) Different.

Some things which could influence the level of re-validation might be:
1 - Are the operating systems the same (Before and after)
2 - Is the hardware different (make, model, manufacturer).
3 - Are there any other functional differences between the old computer and the new computer.
4 -Are you replacing the whole computer, or just the hard drive? If it is just the hard drive, then this should be a lot less work (as I assume the hard drive is compatible with the motherboard, firmware. etc in the computer)
5 - Is the networking configuration changed in any way - are there potential hand-shake or communication issues with the change out?
6 - Etc, etc. etc (there are many things to consider).

I assume that since the hard drive crashed, it may have been older ,and therefore you are not replacing it with the exact same make/model/operating system.

If this is the case, you can review the differences in the old vs. new and review review revision notes/history. Make sure that the software you validation is compatible with the new hard-drive and computer.

After all this assessment, testing can focus on what might change between the old and the new (after you’ve assessed the differences and determined what could change), and then you come up with a re-validaiton plan based on what you know about the changes.

Also, the level of testing and control should be based on what the validated system is/does. If it controls a HPLC, it is pretty easy to redo the whole validation from a functional testing perspective. If it is a SCADA system, then you might need to look into what to re-test more (as SCADA valdiation is very extensive and takes weeks/months).


I agree with JaredCroft, that’s a very good breakdown.

Sometimes places defer to a full Validation because it’s faster (in regards to returning to production) than having to justify a partial Validation.

I would recommend putting that assessment in a disaster recovery plan to make restoring systems in the future faster. The more detailed the plan, the easier it will be to justify partial re-validation and reduce down-time.

Depending on the system, you can restore images rather than reinstalling the software. That would also speed up your restoration.

These are just some things to consider in the event of future hard drive failures, especially if you have other older systems that are nearing end-of-life.

Let me know if you have questions on some of the best practices and I’d be happy to help.

Thanks guys.

It is a like for like change in the sense that the hard drive will be replaced by exactly similar drive, same OS and similar version of software. Since the Software will be reinstalled I am planning to execute a modified IPQ to verify the installation and functioning as there could be potential issues with connectivity and interfaces.

PS: it is a end of the line packing station computer that has the serialization software.