Frequently, "The Vault server could not be contacted...

[jfreidin]

Hi, numerous users here at Carbonite, including myself experience frequent Vault client/server connection failures. This problem is so pervasive that several users have commented that they no longer leave the Vault client running. They close it after every transaction. Some users reported this problem getting worse when we upgraded to 3.5.2.

I'm not saying there aren't problems with our network, although I can't say what those might be. We're running persistent router-to-router tunnels to our remote sites.

Can you modify the client to respond more quickly and retry automatically? My Vault client network settings are:
Use chunked is not checked.
Use Expect is checked.
Request database delta is checked.
Use the system default proxy: No system proxy specified.

First, there is a terribly long 95 second timeout.

Then a pop-up appears, "The Check In/Commit transaction failed. See the message pane for details."

From the message pane:
[6/1/2007 11:03:46 AM] Preparing data to begin transaction
[6/1/2007 11:03:46 AM] Beginning transaction
[6/1/2007 11:05:21 AM] An error occurred while trying to begin a transaction.
[6/1/2007 11:05:21 AM] Transaction failed
[6/1/2007 11:05:21 AM] An exception was encountered during the transaction. Exception: The Vault server could not be contacted to perform the operation. Your network connection to the server may have been interrupted. Please verify your network settings using the Options dialog under the Tools menu in the Vault GUI Client.The underlying connection was closed: An unexpected error occurred on a receive. at VaultClientNetLib.VaultConnection.BeginTx(Int32 nRepID, VaultRequestItem[]& requests, String& strTxID, String comment)
at VaultClientOperationsLib.ClientInstance.Commit(ChangeSetItemColl givenItems, String strChangeSetComment, Boolean keepCheckedOut, Boolean removeLocalCopy, CommitType committype, VaultDateTime dateImport, Int32 nUserIDImport, Int64& nRevID, Int32[]& retBegEndTx)

If I then commit the pending change (thank you for keeping track of what I was doing), it always works, as if the previous connection had dropped, and the client had to reconnect. Can't this be made faster and more automatic?

[6/1/2007 11:19:43 AM] Preparing data to begin transaction
[6/1/2007 11:19:43 AM] Beginning transaction
[6/1/2007 11:19:43 AM] Rename $/Experimental/CarboniteTape/APS_SynchTapeLibraryEnumerations.sql
[6/1/2007 11:19:43 AM] Ending the transaction
[6/1/2007 11:19:44 AM] Transaction completed successfully

Mainly I'm requesting that the Vault client be made more resilient in the face of server connection failures, but any help in diagnosing whatever problem is causing the connection to drop in the first place would be appreciated.

[Beth]

Check your Vault Server Log for what error could be occurring at the time of the connection loss. It could be a timeout of some sort.

Can you post the errors here or the log itself? If you want, you can also just send the log through to support at sourcegear.com (attn: Beth) with a link to this thread.