Sage 100

For those of you who are involved in the backup process of your MAS clients: I have a client on MAS 200 4.50 / Sage 100 Advanced and am in dispute with their outsourced IT guys about backup strategy. MAS is running on a VMWare ESx 5.1 virtual machine (VM) and each VM has the Symantec BackupExec remote agent installed. BackupExec is installed on the host VM and backup media is tape. In MAS 200 even if everyone exits out of the system the App Server will still be running ergo files are in use. This is normal. Also the whole VM is not being backed up just specific folders and the system state. Some questions: 1) Have you found that BackupExec is able to back these open files anyway or does it require an add-on product to accomplish this? 2) Is there another brand of local backup software that you prefer for MAS 200? I normally see BackupExec 3) I mentioned doing image backups of the entire VM on a periodic basis during the day to allow restore to a point in time and this was a new concept to them. Symantec apparently has a VMWare client to allow this but I have no experience with that. I have seen image solutions used like StorageCraft which does sector based backup. Any recommendations on image solutions? I'm meeting with them today and I'll of course talk about complete disaster recovery as well including the EverSafe solution. @Sage100Routine

Hi Alnoor, In response to item 3, I've had very good results with a product called Veeam. http://www.veeam.com/virtual-machine-backup-solution-free.html ~Chris

Backup Exec has an open file add on. The open file agent is more expensive than Backup Exec.

I'm assuming their BackupExec version is 2010 or older, as I believe the newer versions are image-based (VM-centric). 1) There was an Advanced Open File option that would allow the server to use VSS (Volume Shadow Services) to backup any open files. Without that option, I remember it being able to grab all the files except ACTIVATE.PVX, so I would just make a copy of the ACTIVATE.PVX file so that the copy could be backed up. 2) If it's Server 2003 R2, you could just use NTBackup to grab the MAS90 folder as that's VSS-aware. Newer Windows servers use disk/image backups, but if you don't mind getting the whole drive, then that should work as well. 3) Our BDR appliances use StorageCraft as their main backup engine, and that's been solid as a rock for us.

1) appears well answered above 2) we left backup exec years ago and choose to stay out of the on-premise support of IT world 3) We have server replication AND use R1Soft (I think it may be from Idera) as a image solution

Thanks everyone for the great insight! They are on BackupExec 2012 running on Server 2008 R2, but it's not image based by default apparently. I have a quote from Symantec to buy the additional Agent for VMWare/Hyper-V + 1 year Support. @ChrisDiAngelo thanks for the Veeam recommendation. @SteveIwanowski thanks so much for mentioning what is possible w/o the Open File option. Also that the option is VSS aware! And for validating StorageCraft. I'm hearing several thumbs up for that now. @GaryFeldman I have 1 client using R1Soft from Idera for image backup. It works well for them except for the 1st month where they kept complaining every 15 minutes MAS 200 would globally hang. I found R1Soft was configured to backup (probably differentially or incrementally) every 15 minutes (probably not the default setting) so I had IT change the schedule. I do like both the idea of server replication and of images being backed up to the cloud during the day. Or maybe Mark Chinsky's backup appliance / solution. I did tell client yesterday we should opt for something that involves full disaster recovery. I also told them to drop their IT group for the long term once we find a better provider.

This is a perfect application for EverSafe! or EverSafe! lite. It's a complete image based backup of both physical and VM machines and can completely backup all in-use files and fully supports both Microsoft's VSS writers and can fall back to it's own OEM'd Shadowprotect engine. It's injected at the kernel level so it can get at the data at a far lower level than backup exec or even something like Veam and is way faster than Veam. I can backup a typical server (after the initial backup) in less than 90 seconds and with no perceptible performance hit while its happening. Most of my clients take snapshots at least hourly and can 100% verify the backup with proof of bootability on a daily basis. Attached is the screenshot of the typical daily email you get per server. How often are they doing test restores of those tapes? How reliable is their offsite rotation? Given their file only backup, if they had a complete machine meltdown or corruption, how much time would pass before they would be running their business again?

@AlnoorCassim - A very timely topic as a client ran into this issue a couple weeks back. They do a full backup every morning starting at 2 AM. For some very strange reason, BOTH mirrored drives suffered complete failure at the start of the backup. While they were able to restore from the previous night, they lost a complete days work. So now they are exploring adding incremental backups of some sort to cover the period between backups. Probably will never need it, but who thought two drives would blow at once.....

@JeffSchwenk The problem is most backup solutions are 'agentless'. The advantage of not using an agent is you don't need to reboot the server after installation which is a minor inconvenience. The downside is, an agentless backup (like backup exec), needs to traverse the entire tree of files on the server to figure out what's new, what's changed etc in order to do an incremental. If you've ever booted a computer with something like Sugarsync, you'll notice it crawls for quite a while during that process and unless a server is all SSD, it too will suffer noticeable performance hits as the backup program figures out what changed. Thus if your client decides to backup every hour or two, for 10 to 30 minutes that machine could be very sluggish during the backup. And this assumes they are using backup software that can reliably back up in use files like MAS on provideX. Even software that can back up files that are 'in use' can still cause corruptions with MAS because unless you are on SQL, MAS is just a bunch of files. If somebody was committing a transaction where one file got backed up seconds or minutes before others and a posting routine, or a data import occurred across files, you now have an 'out of sync' mas90 database which isn't going to be fun to figure out what's wrong and they may not even notice the problem for weeks or months when certain reports don't balance or tie out.