I have a database which my indexes are in separate filegroups and space (SSD). The SSD which contains indexes is break down, my actual data which are on tables are secured but because of SSD break down, my database is on Recovery Pending mode.

I want to get rid of those filegroups which were on SSD and create a new filegroup on another space, then create my indexes on it, but I can't do anything because of Recovery Pending mode.

I have done these solution:

  • emergency mode, make filegroup offline, make database online
  • partial restore from full backup (my backup is for a month ago) which not worked, database stays in Restoring mode
  • paste deleted mdf from backup into its location, but database still stays in Recovery Pending mode
  • detached database and attach its current filegroups plus deleted filegroups from backup, which not worked because deleted filegroups was from another database and was not accepted by SQL Server

I don't know what to do any more, I have two options:

  1. Make deleted filegroups offline, create new filegroups and create my indexes with new names on them. It is not a permanent way indeed, because there is some deleted filegroups on my database which is not clean, but it help me to back online.
  2. create a script to create new database, new filegroups, new tables and insert current data to them, create stored procedures and triggers and indexes and ... which is difficult to write.

If you have any suggestion I'll be appreciate.

  • May I know, What types of files you have now & the size of database. Somehow I agree with Dan Guzman, But we required .MDF, .dif .trn and taile log backup to restore to complete database . Please share the required information, it will help us to give you the best & doable solution.
    – Jack D
    Aug 10, 2022 at 11:21
  • @JackD I had not ant ,dif or .trn or taile log backup. I see there is no option so I start to create a new database and since I had all of my objects create script I create them, after that wrote a script to insert data from broke database to new one. Everything goes well. Lost mdfs just contained indexes, so I did not lose any data. Thanks for your response Jack. Aug 12, 2022 at 8:09

1 Answer 1


The normal process to recover from a storage failure is to backup the tail of the log (if possible), restore from your last full backup, restore the most recent differential backup (if you do those), and restore subsequent log backups (including the tail).

There is no (at least supported) way to avoid restoring individual filegroups and end up with a healthy database.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.