Network Disconnected - Operation Failed: My reasons to ditch TI

OK, I need some assistance here. I've got some issues with TI 9.1 I need to get resolved or I will have to switch back to Backup Exec. A good backup system should be the LAST thing you have to worry about working correctly.

Currently I have three concerns that are driving me to make this post:

1. Backup speed. I have 4 physical servers, one of which hosts several VMWare images running TI 9.1 as well that all backup to a separate media server with Firewire external drives for storage media. All machines on the network are using dual bonded gigabit network cards in Link Aggregation mode with a Catalyst 2960G switch on the backbone. The average backup time for 50Gigs is a shade over 8 hours with this setup. That seems rather slow. Some servers are faster than others, taking only a couple of hours to backup 100Gigs but some take over 24 hours to do the same.

All Acronis machines are Windows 2003 R2 running the latest TI build.

2. Every single day I get the following error from RANDOM computers in the backup group:

Code:

Event Type:	Error
Event Source:	Acronis True Image Server
Event Category:	(4)
Event ID:	3
Date:		8/16/2007
Time:		12:38:19 AM
User:		N/A
Computer:	CSN-PDC
Description:
The description for Event ID ( 3 ) in Source ( Acronis True Image Server ) cannot be found. The local computer may not have the necessary registry information or message DLL files to display messages from a remote computer. You may be able to use the /AUXSOURCE= flag to retrieve this description; see Help and Support for details. The following information is part of the event: Error writing the file. (0x40003)
    function = "FlushFileBuffers"
    Tag = 0x7CEB2CDC9FB12028
The specified network name is no longer available. (0xFFF0)
    code = FFFFFFFF80070040
    Tag = 0xBD28FDBD64EDB816.

Not the same computer every day... this error message is totally random as far as which server kicked it off.

3. I also get this message EVERY DAY from different servers that do NOT send out the previous error:

Code:

Event Type:	Error
Event Source:	Acronis True Image Server
Event Category:	(7)
Event ID:	4
Date:		8/16/2007
Time:		1:44:47 AM
User:		N/A
Computer:	EXCHANGE01
Description:
The description for Event ID ( 4 ) in Source ( Acronis True Image Server ) cannot be found. The local computer may not have the necessary registry information or message DLL files to display messages from a remote computer. You may be able to use the /AUXSOURCE= flag to retrieve this description; see Help and Support for details. The following information is part of the event: Network disconnected. 
: None.

I have not had a PERFECT backup of all servers since I was using Backup Exec a few months ago. The powers that be wanted me to try Acronis and give them my thoughts, but from a Disaster Recovery Team Leader's viewpoint - it's impossible to feel warm and fuzzy about my data with these continuing errors.

I have searched high and low regarding these errors, found a few hits on here regarding the network disconnected errors, applied the suggested fixes and NO DICE.

One more thing... is there any way to install Acronis or update it without having to restart the machine? We run a 24x7 shop and having to restart servers when applying a patch is overkill. Unless Acronis updates the core OS or installs a boot time driver, a restart shouldn't be necessary.

Sorry for the rant, but I need assistance or Acronis is headed for the dumpster.

 

I don't know if you've tried this already or if your backup tasks even set this way but I would randomly get this error (your first screenshot) until I changed the backup priority from High to Normal. I haven't seen that error now in over a year.

Jon W.

 

bsmith_CSN,

Does this post worry you? It worries me.

https://www.wilderssecurity.com/showthread.php?t=178796#4

I doubt you test restore every image.

 

You should always do a test restore when you first install TI on a machine.
You should do one if you upgrade to a new version.
You should do one if you change your HD.

Do you have to do a test restore on every image, not in my opinion. Once you have determined that you can validate and restore with the recovery CD you have tested out the Linux recovery enviroment and shown it to work with your system. This is where most people come unstuck, they create the archive in Windows and then validate in Windows. Not surprisingly, they think all is well. Then comes restore time and guess what, you use the Linux program, environment and drivers to restore the active partition not the Windows environment you validated in. After knowing the Linux portion works properly with your HW then a validation in Windows is good enough.

My only failure with TI was due to marginal SATA cables which caused the archive to be declared corrupt. Windows had also flagged the problem in the Event Viewer with the specific recommendation to change the cables.

 

seekforever,

I agree. However there are some very concerning posts in that thread and several members have less than full confidence in the ability of TI to restore every image. Xpilot in this post indicates 2 failed restores out of 150 images. I don't like that failure rate.

https://www.wilderssecurity.com/showthread.php?p=1037764#post1037764

 

Remember that Xpilot doesn't validate his images so the only time he catches a fault is on a restore but for the way he works that isn't a major problem.

I disagree with those who don't validate for whatever reason unless they are doing what Xpilot is doing by immediately restoring the image and putting it into service. Even though you confirmed your images worked fine when you installed TI, you could develop a subtle HW fault such as flakey RAM at some time and this could be your downfall if you just stockpile images without validation or restoration.

The validate showed my image to be invalid when I had the SATA cable problem. This is the only time I had this problem but I keep validating anyway.

 

Yes, I've tried setting the priorities and network speeds to different settings in my quest for a complete backup. I didn't see a change at all with any of the settings other than the server load was slightly less.

 

I don't test restore every image, but I do have a set of standby servers that I do a random image restore on once a week and do a full blown disaster recovery excercise once a month using the standby servers. While doing the disaster recovery excercises; Symantec hasn't let me down, BackupPC on Linux hasn't let me down, hell even Microsoft Backup fares better than Acronis at this as Acronis has never succesfully completed a full backup of ALL servers on the same day!

I'm not a set-it and forget-it person and having experience recovering from real physical disasters has instilled a practice of habitual single image testing once a week and a full-on excercise monthly.

Nothing will make your backside pucker like showing up at your datacenter on a military base after a hurricane and find out your whole building is gone - along with your servers!

 

Has this now been solved ? Have the same problem using Ti 10 Home and a PB Netstore 3500. This occurs after 5 hours approximately. The error was reproducable !

 

If by solved you mean "replaced by Backup Exec," then yes

I still have not resolved this issue and the only feedback I've received is what's in this thread.

My data is too important to be trusted to Acronis at this point. If a fix becomes available I will try it and see if it resolves the problem, but I refuse to play with a ticking timebomb.

 

I have solved the error by forcing the backup into 4 GB chunks. Apparantly Acronis TI fails to see the fat32 filesystem on my PB netstore 3500 nas and was writing past the 4 GB filelimit hence the filesystem cache was overflowing and subsequently a network disconnected error was showing in the log. Bad job by Acronis !....

 

If I were using FAT32 drives, this may work but I'm on NTFS.

Just ordered a Powervault 114T with twin LTO-3 drives so the external hard drives are going bye-bye too.