Enhancement: Checkpointing disk tests

Any ideas, thoughts - not necessary related to Hard Disk Sentinel.
toyzrme
Posts: 5
Joined: 2020.11.30. 00:27

Enhancement: Checkpointing disk tests

Post by toyzrme »

Tried your product, liked it, bought 5-license bundle, then discovered the disk testing - LOVE IT!

BUT: twice now I have had failures while running surface tests on brand new large drives (12TB Seagate Ironwolf), both times well past the 75% point. (these were read-write-read tests, so a multi-day affair).

I think this is due to power or hardware issues, not HDSentinel.
BUT, unfortunately, I have to re-start the tests ALL OVER AGAIN.

It would be a huge improvement if you could checkpoint the test(s) on a regular basis:
- let me pick where the log is written (default = Documents, filename=<Disk S/N>+<Timestamp>)
- how often it is checkpointed (by %, time, or GB)
- HDS just writes a new line of text in the log every so often, and flushes/closes the file
Include: timestamp, %/position/progress, # of Good/Damaged/Bad sectors,

Bonus points for including Time & Est., or saving drive map images.

You could also log actual block errors, and let me pick a threshold count where it would abort the whole test if more than X damaged (or bad) blocks were found.

I understand not wanting to re-start a test automatically on reboot, but this way you would at least give me a way to pick up from where it was at the last checkpoint. Of course, if you could read your own checkpoint logs, you could notify me that was the last checkpoint, and offer to restart there, or at least prefill the start point.

UPDATE: Just as I was about to post this, I found your "Automatically save disk test results" option. This is great, *IF* the test completes. If you could record this information regularly, it would be MUCH better. Maybe put each set of reports in a different fold, based on <Disk S/N>+<Test start timestamp>
User avatar
hdsentinel
Site Admin
Posts: 3010
Joined: 2008.07.27. 17:00
Location: Hungary
Contact:

Re: Enhancement: Checkpointing disk tests

Post by hdsentinel »

Thanks for your message and your kind words !

Thanks for the tip, yes, I agree that this would be useful, so probably will be included in a later version.

Yes, when I read the first half of your message, I wanted to recommend the Configuration -> Advanced Options -> Automatically save disk test results to folder option to automatically save results - but yes, this saves only when test is completed.
I do not think a new/different folder (for this purpose) would be useful - but in the above configured folder (if this option enabled) it would be nice to save details about the last started (but not yet completed) tests.

Anyway, you do not need to restart a test from the very beginning:

Please open Disk menu -> Surface test and select the disk drive and test type.
Then, before starting the test, click on the Configuration tab and enable "Limit testing to specific data blocks".
Then it is possible to specify First block / Last block values: for example if you start at block 8000, then only the last 20% of the disk drive will be tested.
(as mentioned in the other topic too).
toyzrme
Posts: 5
Joined: 2020.11.30. 00:27

Re: Enhancement: Checkpointing disk tests

Post by toyzrme »

Actually, I *DO* have to start from the beginning if the system crashed or rebooted on its own - or is there a way to see the results and position just before the crash? If not - start all over again.

Yes, I am using the "Limit testing to specific blocks" now, and testing subsets (segments) of the disk, so that a given test will complete and record results within a reasonable time window, reducing the possibility of a "tested 99% but the system crashed with zero results so I have to start all over again" situation.

I would also argue, if you don't want to KEEP a lot of interim results, at least record the LATEST cumulative results every so often. IOW, every 10% (or whatever), write a complete set of current status/partial results, and erase the previous set. (i.e. write at 10% … then at 20%, write the 20% then erase the 10% report).

*IF* it completes, you erase the last partial report, leaving the complete result just as it does now (if enabled). IF it FAILS, at least I can find roughly where it failed, how it was doing at that point, and can continue from the last written checkpoint (or start a different test if there were lots of drive errors).
User avatar
hdsentinel
Site Admin
Posts: 3010
Joined: 2008.07.27. 17:00
Location: Hungary
Contact:

Re: Enhancement: Checkpointing disk tests

Post by hdsentinel »

I completely agree you, no need to argue ;)

This new feature should work and will be likely designed this way: we would not need a lot of intermediate results, just periodically keeping the last step (eg. once per every 5-10% completed as you wrote) to indicate a checkpoint. And if the test is completed - then we can save the completion reports (like now) and erasing the last checkpoint saved.

I only wrote that there may be no need to configure a different folder for THIS purpose if we already have an option where we can configure a folder for test results.
toyzrme
Posts: 5
Joined: 2020.11.30. 00:27

Re: Enhancement: Checkpointing disk tests

Post by toyzrme »

Oh, ok - my bad!

And I agree, no need for additional folders. I hadn't thought of the "just keep the last successful partial result" idea until partway through my response, eliminating any additional need for organization.

Although … :D LOL
I have a lot of disks on one system - so I could see having a "master folder", with subfolders for each serial number. If I'm going to run tests repeatedly over time, it would be nice organization, but I can certainly do that manually.

BTW, one of my favorite things about your tool is the color coding based on time/ease of operation for each block, rater than just good/bad: I had a new SSD that was light green for the first 80%, but then had a surprising number of darker green blocks in the last 20% (unless this was driven by TRIM operations?). I will be watching that drive over time ...
Post Reply