Hi, just wanted to hear some opinions on whether I should add a second parity drive. My server currently has 4 16TB drives that are about 2 years old, but the drives are full and I recently purchased 2 additional 16TB drives. It is just a media server(my irreplaceable files are backed up more securely in a different way), so I could get all the data back, but 50+ terabytes of data would be a pain to replace.

 

I have seen a lot of people say that you should add a second parity anywhere from 6 to 12 drives. My NAS only has space to add 2 more drives, and 16TB will only last me another couple years. I know the larger the disks in the array, the more risk there is of a failure while rebuilding. Guess I'm just wondering if I'm asking for trouble if I get greedy and only opt for 1 parity drive. Thanks for any thoughts.

If you have daily health reports from the server, regular (typically monthly) parity checks, and pay attention to warning signs by acting promptly, then single parity should serve you fine. If you are the kind of person to let things go as long as they seem to be working fine, and only act when there are obvious signs of trouble, then 2 parity drives could possibly save your bacon.

 

There is no way to accurately predict how and when a drive will fail 100% of the time, many times there are warning signs, and Unraid will try its best to keep you from losing data, but it's up to you to stay on top of your drive health. An Unraid server can continue to serve files seamlessly with as many bad drives as you have parity, but exceed that tolerance by waiting for another failure and you will lose data.

Do you have Notifications setup to alert you immediately by email or other agent as soon as a problem is detected?

@trurl Nope, I run a parity check like once every four to six months and do nothing else.

Just now, Hal9001 said:

@trurl Nope, I run a parity check like once every four to six months and do nothing else.

Unless you regularly check for errors via the Unraid GUI then not having notifications enabled to,advice you of problems as they arise is a recipe for potential data loss. It means that you can have multiple errors creep up on you that you do not notice until they have gone beyond the level of being able to recover without data loss.  One of the features of Unraid is its ability to continue working after a drive failure and if you are not monitoring for this you will quite likely not take appropriate remedial action in time.

What I would like to suggest is that:

 

1-- You setup daily notifications of problem.  The daily check (once setup) will usually only require you to read the message header which only requires a few seconds.

 

2-- You install the 'Parity Check Tuning' plugin to allow you to split your parity check into pieces that run at convenient hours when you are not using your server.

 

3-- You setup a scheduled monthly parity check (using the plugin). These often will show up disk problems while they are still easily manageable.