NS_I_DISK_REBUILD_STARTED (0X400D0056) on Content Server
This event means a disk rebuild started on your Content Server. Usually it's fine, but if it keeps happening, there's a deeper issue with the disk or the server config.
What This Event Means
You're seeing event ID 0X400D0056 with the message "Started rebuilding disk %1 ( %2 ) on Content Server %3". This is from Novell Open Enterprise Server (OES) or a similar NetWare-based system running NSS (Novell Storage Services).
The culprit here is almost always a disk or RAID controller that flagged itself as bad or inconsistent. The server starts rebuilding to fix the data. In most cases, this completes on its own and you can ignore it. But if it keeps popping up or the server slows to a crawl, you've got a real problem.
Here's the troubleshooting flow. Try the easy stuff first. Don't jump to the advanced fixes unless you have to.
30-Second Fix: Check Rebuild Status First
Don't panic. Grab the disk number from the event log. Then run this command on the Content Server:
nss /poolrebuildstatusIf you see a percentage like "45% complete" and it's climbing, let it finish. This can take hours depending on the disk size and server load. Check back in 15 minutes. If the percentage moves, you're fine.
Also check the RAID controller if it's hardware RAID. Most controllers have a management utility like hpacucli, MegaRAID Storage Manager, or omreport. Look for the disk's state. If it says "Rebuilding" with a progress bar, same deal — let it finish.
If the rebuild completes and the event doesn't come back, you're done. Mark it as resolved.
5-Minute Fix: Stop Rebuild Loops
If the rebuild completes but then starts again within a few hours, you've got a loop. This usually means a bad disk or a bad connection.
- Check SMART status on the disk. On Linux-based OES, run:
smartctl -a /dev/sdX | grep -E "Reallocated_Sector|Pending_Sector|Current_Pending"Replace sdX with the actual device. If you see any non-zero values for "Reallocated_Sector_Ct" or "Current_Pending_Sector", the disk is failing. Replace it.
- Check the cable. Loose SATA or SAS cables cause intermittent failures. Reseat the cable on both ends — the disk side and the controller side. I've fixed dozens of rebuild loops just by reseating a cable.
- Check for bad sectors on the NSS pool. Run:
nss /poolscan /nh /skip-mediaThis checks the pool metadata without scanning the media. If it finds errors, note them and then run a full media scan with:
nss /poolscan /nhThis takes time, so schedule it during a maintenance window.
After these checks, if the rebuild keeps looping, move to the advanced fix.
15+ Minute Fix: Replace Disk or Rebuild Pool
This is the "I've tried everything" section. You'll need to take the server offline or at least the pool offline.
Step 1: Confirm the bad disk. If SMART looks clean, run a long self-test:
smartctl -t long /dev/sdXWait for it to finish (check with smartctl -l selftest /dev/sdX). If it fails, replace the disk.
Step 2: Replace the disk. If it's hardware RAID, follow your controller's procedure to remove and replace the failed disk. The RAID controller will rebuild the array automatically. If it's software RAID (like NSS mirror), you'll need to:
- Identify the mirror set:
nss /mirrorlist- Remove the bad disk from the mirror:
nss /mirrorbreak /n=poolname /d=/dev/sdX- Replace the disk with a good one (same size or larger).
- Add the new disk to the mirror:
nss /mirroradd /n=poolname /d=/dev/sdYThe rebuild will start again, but this time it should finish cleanly.
Step 3: If the disk is fine but the pool is corrupted. Rare, but it happens. Back up the pool data first (use NSS backup or rsync). Then destroy and recreate the pool:
nss /pooldestroy /n=poolname
nss /poolcreate /n=poolname /d=/dev/sdXThis is nuclear. Only do this if you have verified backups and the rebuild won't stop.
Step 4: Last resort — check for controller firmware bugs. I've seen old LSI and HP Smart Array controllers have a bug where they keep initiating rebuilds on healthy disks. Update the firmware to the latest version from the vendor. This fixed it for a client on an HP ProLiant DL380 Gen9 with a Smart Array P440ar controller. The firmware was from 2016, updated to 2022 — rebuilds stopped completely.
Common Scenario That Triggers This
This event fires when any disk in a mirror or RAID set goes inconsistent. The most common real-world trigger: a power outage or unexpected shutdown. The server loses power while writing to disk, the disk marks the data as dirty, and on boot, the rebuild kicks off. If the power is stable and the disk is healthy, the rebuild completes and you never see it again. But if the disk has bad sectors or the controller is flaky, you'll see it over and over.
Another trigger: replacing a disk in a RAID array without letting the rebuild finish. Don't pull a disk during a rebuild unless you have to. Wait it out.
When to Call Support
If you've done all this and the rebuild still loops, call Novell/OpenText support or your hardware vendor. Make sure you have the event logs, SMART data, and controller logs ready. They'll ask for them anyway.
One more thing: if the server is still under warranty, don't mess with the disk replacement yourself — let the vendor do it. Otherwise, you might void the warranty.
Was this solution helpful?