Resilver restarts when taking drive offline?

[DCswitch]

I am running RAIDZ2 and I began getting SMART errors on a drive. I did a scrub and then immediately added an additional drive (keeping the problem drive ONLINE) and began the resilver process. It was going along pretty good, about 10% per day. Around the 5th day, it slowed down to about 3% or less per day. After a couple of days went by the process slowed to a crawl. The drive was about 60% resilvered. After reading on a forum that the problem drive could be causing this very slow behavior, I decided to take the problem drive offline. I thought that the resilver process would just continue where it left off and keep on trucking. Instead it started the whole process over (starting back at 0%). It was very disappointing to see. Is this normal and is there a way to do this (in the future) so this doesn't happen again? If not I feel like it's better to always take the problem drive offline whenever resilvering because when a drive has a problem time is of the essence.

 

[rs225]

In your case, raidz2 and a fresh scrub, offlining first would have been fine.

But in other cases, particularly when the user doesn't have enough bays/ports and removes the bad drive, things can get very nasty if they have offlined a good drive by accident. In the case of raidz1, if the bad drive is totally dead, the pool is finished because it is now impossible to import. Because of those accidents that are impossible to see coming, it is best to leave a drive online for resilver until it is obvious it must go offline, and you're confident you aren't going to make a mistake either physically or in the GUI.

 

[DCswitch]

I appreciate your input. But is it normal for the whole process to start over again from scratch? Is there a way to do it so that doesn't happen and it can pickup where it left off?