I had some intense moments with a StoreOnce store not coming online anymore. I'm posting my fix here, maybe it helps others.
We had an issue with a software based StoreOnce store. After one disk of the RAID array failed, the machine crashed (it shouldn't have done that ofcourse...) and now the store wouldn't come online anymore. Btw the OS is on the same RAID array which is still booting, data was still available and in tact. We are running DP on 2008R2 sp1.
V:\>storeoncesoftware --list_stores
Store Name: <Store Name>
Store Id: 1
Store Description: n/a
Store Status: error
Store Autostartable: ON
Store Readonly: OFF
User Data Stored: 7489 GB
Store Data Size: 1176 GB
Deduplication Ratio: 6.4 : 1
Listing of the stores succeeded.
V:\>storeoncesoftware --start_store --name=<Store_Name>
Failed to start the store B2D_Remote.
In one of the store-directories I found a s.bad_integrity file. The contents of that file:
Exception [HousekeepingException]: thrown in (src\dedupe\core\CONISTON\src\corededupe\store\CStoreHK.cpp:719 dedupe::store::CStoreHouseKeeper::processMessage)
Error (9) Bad file descriptor
After moving that file elsewhere and also move the xml file in s_tasks.hk, the store starts but after a few seconds it stops again and the s.bad_integrity file is there again. After investigating some more it seemed the store was actually 'scrubbing' retired media, hence the term HouseKeeping in the error file I assume by now. In the directory containing the s.bad_integrity file there was a folder 'retired'. That seemed to be the culprit, renaming that to retired_OLD made everything work again. After that I issued omnimm -delete_unprotected_media <libraryname> to start the scrubbing process again. After that command the retired folders are back, but when the housekeeping finished they were al gone. I assume it holds data for the retired blocks / housekeeping mechanism. It's been running without issues ever since.
Hope it helps others as well.