Re: [Veritas-bu] NBU cross-site
I agree with Ed, I have media servers scattered around the entire country without any issues. On Fri, Oct 3, 2008 at 10:19 AM, Ed Wilts [EMAIL PROTECTED] wrote: On Thu, Oct 2, 2008 at 3:56 PM, spaldam [EMAIL PROTECTED] wrote: Second, I wouldn't run master/media servers across a WAN; especially since NOM has the ability to manage multiple Master servers from a centralized console. If you lose the WAN, you done and all you backups fail. I disagree - we have media servers scattered in multiple locations connected to our main master and it's been working fine for a long time. Simply put, don't lose the WAN - we have redundant paths to all of our locations with media servers. If we lose WAN connectivity to our remote offices, they're in a world of hurt anyway and backups are the least of their problems. -- .../Ed Ed Wilts, RHCE, BCFP, BCSD, SCSP, SCSE [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] NBU cross-site
On Tue, Oct 7, 2008 at 5:04 PM, Clausen, Matt R [EQ] [EMAIL PROTECTED] wrote: I'm curious though…. From a Disaster Recovery perspective, are you replicating a master across sites to avoid the master server being the single point of failure? If so, what are you using to do it? I've got a master replicating to another site. The catalog is going over there (the server is already a media server), and it's got the right hardware (Decru encryption appliance in the same cluster, same tape drives, etc)., but I've never tested a recovery. Getting the data replicating was the first phase of our DR implementation. For now, I've got the data and Symantec's phone number. One of these days, we'll finish the job and make sure we've actually got a working recovery model. .../Ed -- *From:* [EMAIL PROTECTED] [mailto: [EMAIL PROTECTED] *On Behalf Of *Andrew White *Sent:* Tuesday, October 07, 2008 5:44 PM *To:* Ed Wilts *Cc:* VERITAS-BU@mailman.eng.auburn.edu *Subject:* Re: [Veritas-bu] NBU cross-site I agree with Ed, I have media servers scattered around the entire country without any issues. On Fri, Oct 3, 2008 at 10:19 AM, Ed Wilts [EMAIL PROTECTED] wrote: On Thu, Oct 2, 2008 at 3:56 PM, spaldam [EMAIL PROTECTED] wrote: -- Ed Wilts, Mounds View, MN, USA [EMAIL PROTECTED] If I've helped you, please make a donation to my favorite charity at http://firstgiving.com/edwilts I disagree - we have media servers scattered in multiple locations connected to our main master and it's been working fine for a long time. Simply put, don't lose the WAN - we have redundant paths to all of our locations with media servers. If we lose WAN connectivity to our remote offices, they're in a world of hurt anyway and backups are the least of their problems. -- .../Ed Ed Wilts, RHCE, BCFP, BCSD, SCSP, SCSE [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] NBU cross-site
Howdy, Likewise (depending on the environment) we have clustered masters locally and the replicated to another site as well or for the non-highly available we just have a single master and deal with a 4 hour outage to rebuild the machine. Multiple masters is just a pain in the backside. On Wed, Oct 8, 2008 at 9:04 AM, Clausen, Matt R [EQ] [EMAIL PROTECTED] wrote: I'm curious though…. From a Disaster Recovery perspective, are you replicating a master across sites to avoid the master server being the single point of failure? If so, what are you using to do it? -- *From:* [EMAIL PROTECTED] [mailto: [EMAIL PROTECTED] *On Behalf Of *Andrew White *Sent:* Tuesday, October 07, 2008 5:44 PM *To:* Ed Wilts *Cc:* VERITAS-BU@mailman.eng.auburn.edu *Subject:* Re: [Veritas-bu] NBU cross-site I agree with Ed, I have media servers scattered around the entire country without any issues. On Fri, Oct 3, 2008 at 10:19 AM, Ed Wilts [EMAIL PROTECTED] wrote: On Thu, Oct 2, 2008 at 3:56 PM, spaldam [EMAIL PROTECTED] wrote: Second, I wouldn't run master/media servers across a WAN; especially since NOM has the ability to manage multiple Master servers from a centralized console. If you lose the WAN, you done and all you backups fail. I disagree - we have media servers scattered in multiple locations connected to our main master and it's been working fine for a long time. Simply put, don't lose the WAN - we have redundant paths to all of our locations with media servers. If we lose WAN connectivity to our remote offices, they're in a world of hurt anyway and backups are the least of their problems. -- .../Ed Ed Wilts, RHCE, BCFP, BCSD, SCSP, SCSE [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] NBU cross-site
hi its probably about 10 TB per site per night, 200 clients each site, oracle is the main data source, most will be SAN based. There are a small number (20) apps at RPO/RTO= near zero, which are cross site clustered, most other HA/DR is 4hrs, and backup needs to be one of the HA apps because of the potential requirements to recover other apps, also that apps might move from prod one site to prod at the other site and we need to track backups. Pipe is 2 x 10GBE, business is telecoms. Essentially i want these two sites to act as one site. regards mark On 5 Oct 2008, at 02:00, oersted wrote: with 5.x you could do this easily without clustering...but with 6.5 what is the amount of data at eaach site? how many clients at each site? what are critical apps? where is data stored? local? SAN? what does HA/DR SLA entail? if you HA/DR/BC the apps do you have to HA/DR the backup? what is pipe between sites? what is your business? Mark Steel wrote: Hi I need to design and build a NBU config for a 'virtual datacentre' with two sites, both with production hosts. I also need to keep systems to a minimum, and I need HA/DR. I was thinking about cross-site clustered master/media server as the start. It would be Solaris platform so could run functions (such as the master server) in a zone for failover, and media server if required as separate zones. Anyone have any experience of this ? I guess its a balance between complexity of setup with trying to keep config to a minimum - I mean I could have simple local HA / cluster with remote failover an then do that both ways, but that might end up as 6 servers for initial config. Thanks Mark ___ Veritas-bu maillist - Veritas-bu at mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu + -- |This was sent by [EMAIL PROTECTED] via Backup Central. |Forward SPAM to [EMAIL PROTECTED] + -- ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
[Veritas-bu] NBU cross-site
with 5.x you could do this easily without clustering...but with 6.5 what is the amount of data at eaach site? how many clients at each site? what are critical apps? where is data stored? local? SAN? what does HA/DR SLA entail? if you HA/DR/BC the apps do you have to HA/DR the backup? what is pipe between sites? what is your business? Mark Steel wrote: Hi I need to design and build a NBU config for a 'virtual datacentre' with two sites, both with production hosts. I also need to keep systems to a minimum, and I need HA/DR. I was thinking about cross-site clustered master/media server as the start. It would be Solaris platform so could run functions (such as the master server) in a zone for failover, and media server if required as separate zones. Anyone have any experience of this ? I guess its a balance between complexity of setup with trying to keep config to a minimum - I mean I could have simple local HA / cluster with remote failover an then do that both ways, but that might end up as 6 servers for initial config. Thanks Mark ___ Veritas-bu maillist - Veritas-bu at mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu +-- |This was sent by [EMAIL PROTECTED] via Backup Central. |Forward SPAM to [EMAIL PROTECTED] +-- ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] NBU cross-site
On Thu, Oct 2, 2008 at 3:43 PM, Mark Steel [EMAIL PROTECTED] wrote: It would be Solaris platform so could run functions (such as the master server) in a zone for failover, and media server if required as separate zones. A media server in a non-global zone is not supported. I think that's probably true for a master as well but I haven't looked that up recently. .../Ed Ed Wilts, RHCE, BCFP, BCSD, SCSP, SCSE [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
[Veritas-bu] NBU cross-site
Hi I need to design and build a NBU config for a 'virtual datacentre' with two sites, both with production hosts. I also need to keep systems to a minimum, and I need HA/DR. I was thinking about cross-site clustered master/media server as the start. It would be Solaris platform so could run functions (such as the master server) in a zone for failover, and media server if required as separate zones. Anyone have any experience of this ? I guess its a balance between complexity of setup with trying to keep config to a minimum - I mean I could have simple local HA / cluster with remote failover an then do that both ways, but that might end up as 6 servers for initial config. Thanks Mark ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] NBU cross-site
I agree. Though it would be more expensive (for licenses), I also recommend a Master at each site. If you do the remote method, you will eventually lose the link to the Master and you won't be having much fun. Rusty Major, MCSE, BCFP, VCS ▪ Sr. Storage Engineer ▪ SunGard Availability Services ▪ 757 N. Eldridge Suite 200, Houston TX 77079 ▪ 281-584-4693 Keeping People and Information Connected® ▪ http://availability.sungard.com/ P Think before you print CONFIDENTIALITY: This e-mail (including any attachments) may contain confidential, proprietary and privileged information, and unauthorized disclosure or use is prohibited. If you received this e-mail in error, please notify the sender and delete this e-mail from your system. spaldam [EMAIL PROTECTED] Sent by: [EMAIL PROTECTED] 10/02/2008 04:16 PM Please respond to VERITAS-BU@mailman.eng.auburn.edu To VERITAS-BU@mailman.eng.auburn.edu cc Subject [Veritas-bu] NBU cross-site First, you cannot run NetBackup servers in a Zone. The UNIX admin here already tried that and the install script even said it wasn't supported. Second, I wouldn't run master/media servers across a WAN; especially since NOM has the ability to manage multiple Master servers from a centralized console. If you lose the WAN, you done and all you backups fail. Third, with new de-duplication and replication products that are available today, they offer a great solution for doing your own self insurance with each site having a replicated copy of the other. I'm going to be using Quantum’s DXi units here soon to do just that. +-- |This was sent by [EMAIL PROTECTED] via Backup Central. |Forward SPAM to [EMAIL PROTECTED] +-- ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
[Veritas-bu] NBU cross-site
If you can only aford one master, I wouldn't cluster it across the WAN. If your link goes down and both nodes start acting as the primary node, you'll never get them back in sync without blowing away something valuable on one or the other. Also, I don't think NetBackup masters are supported in an Active/Active envorment. Pulse I'm not sure how that will work for the catalog information. It would have to be replicated between sites somehow anyway. Keep your master at once site, and the media server at the other site, then send your tapes from the one to the other for DR purposes. If you lose the master, youll have to rebuilt it at the other site. +-- |This was sent by [EMAIL PROTECTED] via Backup Central. |Forward SPAM to [EMAIL PROTECTED] +-- ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] NBU cross-site
On Thu, Oct 2, 2008 at 3:56 PM, spaldam [EMAIL PROTECTED]wrote: Second, I wouldn't run master/media servers across a WAN; especially since NOM has the ability to manage multiple Master servers from a centralized console. If you lose the WAN, you done and all you backups fail. I disagree - we have media servers scattered in multiple locations connected to our main master and it's been working fine for a long time. Simply put, don't lose the WAN - we have redundant paths to all of our locations with media servers. If we lose WAN connectivity to our remote offices, they're in a world of hurt anyway and backups are the least of their problems. -- .../Ed Ed Wilts, RHCE, BCFP, BCSD, SCSP, SCSE [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
