Tuesday, December 03, 2019

Recovery Complete: Now Don't Do That EVER Again



Come Undone  indeed.

So it is now over 5 days and a similar number of sleepless nights since a set of freak circumstances rendered all Marcus' in country data and backups faulty.

I know,  what are the chances?

But as of  20191203 04:21

Everything Is Restored


- 50 TB moved, copied or restored

- NAS is rebuilt with RAID5 and Qtier SSD tiering

- Main server is recovered and now uses multiple (old, but still excellent) HGST Enterprise Hard Disks.   They just don't make they like that anymore

- The f****ing ReFS Disk format upgrade that caused all the shit in the first place is no more.   Microsoft: you really let me down here.  Fancy making a disk format that turns to RAW when you upgrade your system, and there is no way to recovery your data on that upgraded system.   Unflippin believable

- I'm back to using NTFS formatted disks



- I'm using Microsoft Storage Services and NVMe disks and I'm now using a multi card, multi disk array solely based on NVMe disk technology.





- The new main workstation disk array is only made possible thanks to Dramatic price cuts at Amazon.  The price of the 1TB hard disk has fallen from 230+ GBP to only 155 GBP.  That is utterly astonishing.

I've literally been waiting for 12 months, saving money, now with these lower prices the final upgrades have been made.




- Thanks to Total Commander, that retro copying tool that I still use daily,  I was able to copy the recovery files to the new destination and keep the creation times.  So all my data still sorts correctly.



- I took the Black Friday upgrade to SyncBackPro  (at the time I was desperate, but looking back it was a decent priced upgrade)

This new version has some good features to enable me to keep my source and target data stores in sync.  Example Fast Backup when it stores file details in a private database to speed up file comparison times.  Just one out of about 1000 impressive features.  Really excellent File Sychroniser.

- I cross compared Total Commander copy logs with SyncBack Pro and found to my alarm that whilst all files were copied some metadata was incorrect.   By Metadata I mean NTFS attributes to the file and also some instances of creation date corruption.

Luckily SyncbackPro has the ability to correct the destination copies with respect to a good source, so this last step took another 12 hours or so.


And Relax
It is humorous but perhaps just a little weird.  When my master computer is down, I become rather anxious.  And when it is correspondingly not just backup, but upgraded, and now just flying along, well, I am just elated.

It does not quite make up for the initial stupidity caused by

a) too much cost cutting and optimisation
b) too many parallel upgrades   (impatience)








We want your Soul