Summary:
Copying the FSA archived items to another place from a network path, the placeholders change back to original icons but the recalled files seems empty, corrupt
The issue:
The customer wants to copy an archived file or files for example to his personal Windows desktop from a network directory. The files are recalled, placeholders change back to the original icons on the network path, the file are copied to the desktop, but when he wants to open any item, for example Excel or Word files, the application says: the file is corrupt. He tries to open same file with Notepad and they seems empty.
Then he tries to open same recalled file from the network directory the issue is exactly same.
When he opens another archived file (looks placeholder), for example Excel (xls) directly from the network with double click it opens successfully.
The environment:
Symantec Enterprise Vault 10.0.4 HF3 on Windows 2008 R2 servers:
-one server for Exchange mailbox archiving
-one server for Exchange journaling archiving
-one server for file servers archiving
-the vault store partition are locally on NTFS volumes
-shared MS SQL 2008 R2 cluster for EV databases
The file servers are:
Microsoft Cluster Server is with two physical nodes, the virtual file servers are the cluster resources. The shared disks are accessed via fiber channel on the storages. The volumes are NTFS. The shared directories on the virtual file servers.
Troubleshooting steps:
I set verbose Dtrace for the Placeholder service on the active cluster node and monitored what happened when a file is recalled with a “double click” opening or a “copy somewhere” opening. I did not experience any difference.
I checked what is in the archives with archive search, I restored an item, I could open successfully, so archive is good (so far).
I tried to recreate placeholders on recalled files with FSAutility –c option and retry again, didn’t solve.
I tried to restore the original files to the original places with FSAutility –t option and rearchive again, didn’t solve.
I restart the EV services on both cluster nodes, didn’t help.
I reinstalled the FSA agents via VAC, didn’t help
I rebooted the server file cluster nodes, didn’t help.
I rebooted the Enterprise Vault file archiving server, didn’t help.
I’ve no other idea, what can be the solution?
I have found and interesting thing:
I tried to open a previously recalled and “corrupt” file locally logged on to the active physical cluster node with the Notepad, the file CONTAINS VALID data. After that opening, the customer CAN open this file successfully via the network path too!
I monitored this phenomenon with Dtrace but it nothing was logged.
