Networker

Re: [Networker] SLAs and backup cost estimates

2008-05-16 13:47:35
Subject: Re: [Networker] SLAs and backup cost estimates
From: Stan Horwitz <stan AT TEMPLE DOT EDU>
To: NETWORKER AT LISTSERV.TEMPLE DOT EDU
Date: Fri, 16 May 2008 13:44:20 -0400
On May 15, 2008, at 9:45 AM, Fazil.Saiyed AT anixter DOT com wrote:


Hello,
Care to share your success, what have you done differently, tape management, media rotation, scheduling, procedures, implementation, hardware to get where you are now. There is growing dissatisfaction among other networker admins in regards to Networkers high hands on approach, scripting, deficiencies in media management, AD restore capabilities compared to new backup software, bare metal restores,cloning,staging,device management, reporting etc. Perhaps the forum can consider having resident expert in each major area i,e cloning, san & publishing relevant solution that list members can draw upon at will, i.e NDMP backups, Cloning, sricipting. This are just the thoughts, i mean i am sure plenty of admins may feel Networker is up to the task,but the questions still remains is it ?

We do DR tests several times a year on different servers. These tests are audited by our data security people, and in some cases, by an outside firm. Our NetWorker 7.4.1 server handles all those DR tests. We have something like a 98% success rating with our DR tests. The missing 2% are mostly the result of certain system administrators failing to prepare for their DR test properly, such as learning how to use NetWorker's recover utility or due to a problem with the DR server, such as it not being kept up to date.

We are a medium size NetWorker site, with one data zone consisting of three storage nodes and around 300 backup clients and around . Most backup data goes directly to tape (two tape libraries) and some goes first to a disk device (a Sun X4500) then is staged to tape. We do not do any cloning (other than that which is handled by the staging feature). I have staging set up just through the NetWorker console manager, no scripting for it.

We are not set up to do bare metal restores via NetWorker, but I am pushing to get EMC's HomeBase to handle that need for Windows servers. So far though, management is reluctant to pay for that functionality, esp. since HomeBase does not yet support Linux.

Our clients are a mix of Linux, Windows, some Solaris, a bit of Tru64 Unix, and six NDMP devices.

Next week, I plan to upgrade the storage nodes and server to NetWorker 7.4.2. Probably in early July, I intend to promote my one Linux storage node to a server. When that happens, I will have one data zone driven by Solaris 10 with a Solaris 10 x86 (our Sun X4500) storage node both sending data to Sony S-AIT1 tape on a Sony PetaSite. The new server will send its data to LTO-3 tape in a Qualstar tape library.

Our largest tape library (PetaSite) and our NetWorker server sit in a separate building from where our primary data center is located, so we do not worry about shipping most clients' backups off site. The only data we ship off site is for our payroll, grading, and other drop dead mission critical servers, and that adds up to an average of three LTO-3 tapes per day.

I don't know what you mean by "device management." One area where I am very deficient is in reporting, but I am looking to improve on that area. I do ship off daily bootstraps to a gmail address, and I also send off a daily "mminfo -v" for the one client that we would need to recover first in the event of a major catastrophe (payroll). I have also disabled index saves on all my save groups and once a day, I run a "savegrp -O" for all the clients.

Updating from 7.2.1 to 7.4 brought with it a lot of pain, but the update from 7.4 to 7.4.1 fixed most of the problems I had. The only time I ever saw a recover fail was back in July when I personally tried recovering some data for a colleague and we had the dumb bad luck of having that data reside on a defective tape cartridge. Fortunately, I was able to pull the data off the previous day's backup, which was on a different tape.

To sign off this list, send email to listserv AT listserv.temple DOT edu and type 
"signoff networker" in the body of the email. Please write to networker-request 
AT listserv.temple DOT edu if you have any problems with this list. You can access the 
archives at http://listserv.temple.edu/archives/networker.html or
via RSS at http://listserv.temple.edu/cgi-bin/wa?RSS&L=NETWORKER