Re: [Veritas-bu] Is the veritas-bu list still alive?
* WEAVER, Simon (external) [2017-05-30 05:38]: > Is there a new list people are actively subscribed on? > Simon That was my thought as well. There seems to be a general trend toward other social media outlets for a lot of things instead of mailing lists. Maybe NBU is just so easy now, nobody needs to ask any questions about it anymore? ;-) -- David Rock da...@graniteweb.com ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] Archive list
* Kevin Holtz [2012-03-09 07:47]: > All is true but I wouldn't say it's something NBU was not designed to > do. It was made difficult for a reason rather than easy due to the > nature of the outcome e.g. Archiving. This is very common practice > for DBA's. Yes, this all needs to be scripted, blat etc. and you will > need admins to help if you don't have access to the local systems. By "not designed to do" I specifically mean that NBU is not designed to have the backup admins be responsible for running archive jobs. Archives are inherently dangerous and as such, only the owner of the files being archived is really intended to be the one to determine of archiving is ok. Backup admins are generally in the business of preserving data for recovery, not removal of files from systems. -- David Rock da...@graniteweb.com pgpGwem0CmOpv.pgp Description: PGP signature ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] Archive list
* Dennis Peacock [2012-03-08 07:14]: > Patrick, > I'm not a windows guy, so I'm trying to see what I can do without > having to find someone that can write a "windows script" to do this. > > What I'd prefer to be done is for me to setup a backup policy in NBU > that is an Archive type policy. Point that policy to > D:\Archive_Data\LZ and schedule the backup to run every night at 22:00 > hours. When the backup finishes successfully, it would remove the > files and then run the bpflist or whatever to generate a list of what > was backed up and email that to the customer. > > That's what I'd prefer to do, but I'm lacking in knowledge in this > windows stuff as I've been living in the Unix/Linux world for over 30 > years. [Wink] Well, First of all, there is no such thing as an "Archive" Policy type. The closest thing you would have is a Policy type of "MS-Windows" with a schedule type of "User Archive". Archives are designed to be User initiated, not Scheduled from the Master Server. As for notifications, especially because of the nature of Archives, that is something that is going to pretty much _have_ to be a Windows script on the client system. You would have the logic in the clientside script to manage all of that. There are a couple likely issues: 1. The Windows server doesn't have an MTA to send email with (typically there is no sendmail or equivalent on Windows, so that would need to be added) 2. It's pretty much guaranteed that this won't be something that will be managed by you, it will be the Windows admins (although you would be responsible for making the User Archive schedule available to them). You _might_ be able to come up with some insane scheme like using a scheduled backup that touches the client system, causes the clientside script to run, and then uses bpend_notify once it's done to do all the other heavy lifting. That still won't get you away from Windows scripting. Basically, what you are trying to do is something that NBU is not meant to do. The expected behavior for Archives is that they are managed by the client admin, not the backup admin. Archive jobs get a filelist from the user, run the archive, and delete the files when they are done, so the user would already know what files were archived because they would have supplied the list to the command in the first place. You would, at least, be able to query the backup info from the master the following day and report what was backed up, but that's about it. -- David Rock da...@graniteweb.com pgpoF2yhXmc83.pgp Description: PGP signature ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] exclude_list - exclude core file but not core directory
* Anurag Sharma [2011-08-29 10:27]: > > Dave, > > To exclude all files with a given name, regardless of their directory > path, just enter the name without a preceding slash. For example: > "example" rather than "/example" > > Here is an example *exclude_list* > --- > # this is a comment line /tmp/example/dir/all /tmp/example/dir2only/ /usr/home/*/all_tmp /*/all_temp core The issue that I'm having is not that I need to find all instances of core regardless of path, but rather that the exclude_list is picking up both files AND directories named core. What I'm trying to do is exclude all instances of core _files_ but still back up all instances of core _directories_. -- David Rock da...@graniteweb.com ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
[Veritas-bu] exclude_list - exclude core file but not core directory
Hello, I'm dealing with exclude_list for the umteenth time, and stumbled into a classic problem. The examples in the docs have always shown excluding "core", which excludes all core files AND core directories. Because of how you define things, it is possible to exclude just directories by appending / (e.g., core/), but there is no corresponding way that I can find to exclude ONLY files named core. I can more or less accomplish what I need to by using an include_list that contains "core/", but this will obviously add processing overhead because it's going to build the initial list, drop excluded stuff, then go back through all the excluded directories and look for all directories named "core". I especially don't want the include list to go back and re-add any core directories under other directories that I have also excluded. For example: exclude_list: core /dev /sys /mnt/auto /var/mqm/ /u[0-9]*/ If I have an include_list that looks like this: core/ It will pick up any directories named core under: /dev /sys /mnt/auto /var/mqm/ /u[0-9]*/ Which I do NOT want it to do. Does anyone know if it's possible to define JUST files named core in the exclude_list, so that an include_list is not necessary? If an include_list is the only way to do it, is it possible to avoid the re-adding under the directories I _do_ want to exclude? Thanks. -- David Rock dave...@graniteweb.com ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] Firewall configuration setup
* rcarlisle <[EMAIL PROTECTED]> [2007-10-12 07:25]: > Just an update on this one. My client had the firewall set up wrong, they > had inside to outside connections rather than inside to inside connections > set up...once that was fixed, we were great! Yeah, that _might_ do it ;-) -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] NBU 4.5FP_6 error bpprm from Solaris 10 client - cannot open
* hastonian <[EMAIL PROTECTED]> [2007-10-10 18:17]: > > Hello, any comments would be welcome. > > Even when using an exclude file, we cannot get this error message to go away. > > The NetBackup server is a Solaris 9 box and has a Solaris 10 client. > A full backup of "/" returns a single error on /etc/mnttab - cannot open file > . > > When adding '/etc/mnttab' to the exclude file, still the errors occurs. > > The exclude file has /etc/mnttab' but the error returns "/etc/mnttab/" - not > the extra training slash. > > Any ideas? Thanks! -J [EMAIL PROTECTED] If you are using 4.5FP6 on a Solaris 10 client, you are probably well outside of what is expected to work (4.5 is dead, and was long dead before Solaris 10 ever existed). That said, mayby putting in what is complained about in the exclude file will do the trick (i.e. if the error is "/etc/mnttab/" then try "/etc/mnttab", not "/etc/mnttab"). Sorry if that's not particularly helpful. Are you getting a status 1 because of the file? Considering how far out of a supported config this is, you may just have to live with the status 1. I would, however, be concerened that other data may be at risk on the client even though it's not generating errors. -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] NBU Vaulting: Status 50
* Wayne T Smith <[EMAIL PROTECTED]> [2007-10-12 14:19]: > I get a few of these (v5.1MP6), mostly during high-load situations. > Inevitably, the status 50 Duplicate job has this in (the middle of) its log: > > Warning bpduplicate ... failure logging message to client ... in log > ...: can't connect to client (58) We found that to be a problem where the TCP_TIME_WAIT on the master server existed for too many sockets open. We were able to address it by lowering the wait interval so that the sockets cleaned up faster. You do have to be a bit careful with this, however. Setting it really low could have negative impacts. -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] Firewall configuration setup
* rcarlisle <[EMAIL PROTECTED]> [2007-09-25 19:39]: > > Can someone please help me with the following configuration: > > Client - NBU 5.1 > Master/Media - NBU 6.5 > > Port 13724 VNETD is opened both ways. What needs to be set up in NBU on > Master server and Client...global attributes, firewall settings...reserved > portsI still can't get the master server to see the client in hosts > attributes. There's pretty extensive documentation in the manuals for this (more than most people should _ever_ read), but the minimum you need is the following: Server -> Client 13782(bpcd) Client -> Server 13724(vnetd) Client -> Server 13720(bprd) - but only if you need client initiated stuff. In order for vnetd to work properly, you need to add the no connectback option in the client properties on the master server only. You can get to the client settings using bpclient, too. vnetd Server -> Client is pretty useless. There are some other things you _could_ open up, but that should get you where you need to be. -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] NetBackup 6.5 VTL direct to tape
* Paul Keating <[EMAIL PROTECTED]> [2007-08-30 11:09]: > OUCH > > So you mean to tell me that you only have as many virtual drives as you > have physical drives? > > That wouldn't have even made it onto my "for consideration" list, when I > was evaluating VTLs. No, it's a 1:1 for _barcodes_, not drives. -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] Netbackup 5.1 and VMware workstation
* Gary sperano <[EMAIL PROTECTED]> [2007-08-04 17:09]: > > Looking for assistance. Situation: Windows 2003 Server running > Netbackup 5.1MP6. On this server I am also running VMware workstation > 5.5.4 From SUN's web site, I have been using a Solaris 10 X86 > appliance. I am now looking to backup my SUN X86 /export/home > directory from my physical W2K3 server,where NB is installed. > Problem: Can not make a connection to my SUN OS. I have done some > troubleshooting, such as bpclntcmd on the client and on the server > (output below). Also below, will be the ipconfig(W2K3) and > ifconfig(SUN). My Sun server is obtaing an IP address of > 192.168.153.128. The NIC setting for SUN, via VMware, is is to > NAT(used to share the host IP address). When I issuse bpclntcmd -hn > solaris10-vmware, it respons withan ip address of 192.168.153.1; NOT > .128 as I would have hoped for. but when I try a backup, I get error > code 54. You are likely not going to be able to get this to work using NAT with VMware. The server doesn't have any way to get to the guest. You will need to either set up the guest with a bridged connection (so it can get na IP the server _can_ find) or search for a way to get the VMware host to direct traffic to the NATed address (yuck). -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] Maximum number of policies
* Curtis Preston <[EMAIL PROTECTED]> [2007-07-24 18:21]: > You're saying if you manually re-run a backup via the activity monitor > during the daytime, it will fail with a 196? I didn't realize that. That would make sense because re-running is functionally similar to rescheduling a job, which takes the windows into account. You _can_ manually submit a job from the policy listing and that would run without the 196 (unless something else caused it to fail and retry), but that would be the entire policy, not just the specific drive stream. -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] Maximum number of policies
* Justin Piszcz <[EMAIL PROTECTED]> [2007-07-23 17:43]: > Perhaps, I just hope they never had to change a storage unit for all those > 4,000+ polices :) Nah, that's what the commandline is for :-) -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] retentions
* Brandon Zermeno <[EMAIL PROTECTED]> [2007-07-19 15:50]: > We had an issue where was server was down for over 2 weeks and the last > full backups expired. The retention for this environment is 2 weeks. > After the tapes expired the App team decided they wanted to restore. Now > management wants Veritas to extend the retention period automatically if > a full is not successfully run. I do not know of any company that has > this capability so I am looking for ideas. I do not think it would be > possible for a person to manually track all the servers and their last > full and then manually bpexpdate the images. I have pushed back with the > question of why was this server allowed to be down for 2 weeks but > nobody is answering. If the policy was a two week retention, and the app guys knew that, it sounds like they are out of luck. Would the response have been any different if they were trying to get an old file? At a minimum, a review of your retentions for that system (and probably the rest of them, too) is in order. You _could_ do it with some scripting, but you are asking for headaches. There are a lot of questions you need to answer in order to pull this off, like; are you comparing against a flat timespan, or will it change per policy/client/whatever? how soon do you want to extend the retentions, immediately following a failure, 1 day, 1 week, etc? How far do you extend it? What about notifications? Is blindly doing this going to deplete your tapes, causing more issues than it fixes? There is no substitute for a person actually looking at what's going on in the environment. It sounds like what you really need is to look into better ways of reporting on what's happening, and not just the backup admins. The managers and admins of the systems being backed up should be watching, too. That can be homegrown or third-party. The data is there, it just depends on how much effort you want to (or can) put into it. -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] same job keeps hanging.
* Aaron Mills <[EMAIL PROTECTED]> [2007-07-09 16:39]: > Hi all, > > I'm hoping someone's seen this before. I'm running 5.1MP6 w/ AIT3 - I've > got a ~126GB backup that kicks off weekly, but hangs within a few hours > every time - the error I get is always "media manager terminated by > parent process" but the logs don't seem to show anything odd. No other > backups hang like this. This is also the only job that runs on the > server itself. When you say "runs on the server itself", what do you actually mean? We say an odd timeout that always happened at the same time into the backup, but the specific circumstances were: 1. a bpbackup command running on a client system 2. client on the other side of a firewall What was happening in our case was the backup would start, one hour into the backup, the firewall would decide since it didn't see any traffic coming from the client to the master server, it would drop the entry in the state table. Then, one hour later, the client would try to send a keepalive packet through the now-defunct connection, fail, retry several times, and then finally give up and die, taking the backup with it. This may not be anything like what you are dealing with, but it is a pretty good example of how things other than NBU can cause weird things to happen and make it look like NBU is the cause. Does your job always die at the same time, or does it vary from attempt to attempt? -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] Aptare and NBU not reporting properly ...
* [EMAIL PROTECTED] <[EMAIL PROTECTED]> [2007-06-15 16:04]: > Greetings; > > I worked with Aptare Tech support, and I believe we have solved the > problem. Solved by installing a new "aptare.jar" file on the master > server. For some reason the old version was not updating the job details > for all jobs, We'll see what happens over the week.end. Yeah, you have to make sure all your master servers have the current aptare.jar files. They have a process that automatically updates the master server client file, but your environment (firewalling, etc) may not allow it. For example, our environment would update only 5 out of the 7 master servers automatically, leaving 2 for us to do manually. -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] any trainings available for Netbackup software
* sunanda a <[EMAIL PROTECTED]> [2007-05-29 11:24]: > Hi all, > are there any trainings available for Netbackup on > differentplatforms... > If anybody knows please give me details.. I wanted to go for detail training > n netbackup. > Handson and practicle training... There are a number of different places to get training. You can start by visiting the symantec website: http://www.symantec.com/enterprise/training/index.jsp -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] Starting an Archive backup from a policy
* Schneider, Matthew J. <[EMAIL PROTECTED]> [2007-05-16 09:19]: > Hello, > > I'm trying to setup a policy to start an Archived backup from the > media server instead of from the client. In other words I would > rather not use a script to start an archived backup from the client in > the case the scheduled task or script fails, I would not find out > about in the Activity Monitor. no, no, NO. User Archive backups can be run only from the client system for a reason. You want the USER to be the one to shoot themselves in the foot, not you. User Archives are bad news, ESPECIALLY scripted ones. All it takes is one filelist entry to be wrong and your whole client system is GONE (backed up, but gone regardless). So, unless you are the admin of that client system, do not go near it. A better option is to understand what you are trying to accomplish and see if there is a better (safer) way to do it. Archive jobs are not really intended to be used as a space-saving technique, they should be used to Archive data (Take static data that is logically related and store it off the system). That is not the same thing. Having said all that, of course there are ways to do it (bpstart_notify on the client for one) but you should really re-evaluate your needs. -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] Start NBU non-root
* Ed Wilts <[EMAIL PROTECTED]> [2007-05-15 21:01]: > One word of caution - if your script isn't absolutely rock solid, you could > potentially set yourself up for a world of hurt. For example, if you allow > apache to run bprestore via sudo and don't properly restrict the source and > target destinations, you could find yourself allowing a user to restore > ~/myownpasswd.file to an arbitrary Unix host and now you have one or more > compromised system. If the user can restore the passwd file on your > NetBackup master, you now have a totally compromised environment since he > can now restore anything to anywhere as well as having full read access to > everything. I couldn't agree with this more. We had a couple commands that we allowed certain users to sudo to that were READ ONLY tools, like bppllist. I don't even like our operators having access to the Activity Monitor through the java GUI because it's not just viewing. NBU has a wretched excuse for a security model. -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] how to enable bpcd port on Windows 2000 server
* nguytom <[EMAIL PROTECTED]> [2007-05-09 14:01]: > > Hi All > > I couldn't telnet to windows 2000 client with bpcd port > > [EMAIL PROTECTED]:/usr/openv/netbackup/logs/bptm]>telnet crdcasbcptest06 bpcd > Trying 162.123.108.62... > Connected to crdcasbcptest06.us.aegon.com. > Escape character is '^]'. > Connection closed by foreign host. > > I checked in service tab, netbackup services client is running, I > stop/restart it again but I couldn't telnet to this window client by bpcd port You can verify that your system is actually listening on the port with: netstat -an and look for LOCALIP:13782 in a LISTENING state You ARE connecting to it, (as evidenced by the "Connected to" statement in your output). The connection is just not STAYING open. Make sure you have forward and reverse lookups on the client for the backup server you are connecting from. Also, what is your network configuration? Do you have multiple NICs on the client? Are there any firewals in between? Make sure your routing is set up correctly, too. -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] excluding a single host
* Jeff Lightner <[EMAIL PROTECTED]> [2007-04-19 13:18]: > Just modify the policy to remove the 5th host. When you're ready to > resume add the 5th host back to the list. > I had heard that a MP or perhaps 6.5 will finally have the ability to simply check a box to temporarily disable backups of a single client in a policy. Not holding my breath, though. Being forced to remove a client to do this is just stupid. -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] Reporting scripts
* Jeff Cleverley <[EMAIL PROTECTED]> [2007-04-17 20:54]:
> Darren,
>
> I actually feel better knowing that there wasn't a real easy solution
> :-) I did some experimenting and came up with a way that seems
> relatively accurate for what I want.
>
> >>
> cd /usr/openv/netbackup/db/images/
>
>
>
> if [ -f /tmp/stats ]
> then
> rm /tmp/stats
> fi
>
>
>
> ls > /tmp/clients
> for CLIENT in `cat /tmp/clients`
> do
> cat $CLIENT/STREAMS |grep -v ^T | awk '{print $2,$4,$8}' |grep -v ^0
> |sed 's/$/ '$CLIENT'/' >> /tmp/stats
> done
>
> <<
The only downside to doing it this way is it will break horribly once
Symantec decides to change the underlying database structure. Combining
several CLI commands is a safer, albiet longer, way to do it. Just
remember that they _do_ like to change things from time to time ;-)
--
David Rock
[EMAIL PROTECTED]
___
Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu
http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] bpdbjobs displays month number only in date fields
* [EMAIL PROTECTED] <[EMAIL PROTECTED]> [2007-04-11 17:33]: > Hi, > > I have recently upgraded a number of Solaris-based NetBackup > installations to version 5.1MP6. Subsequent to the upgrades, the > "Started" and "Ended" columns of the "bpdbjobs" output display the month > number only. There is a known issue with 5.1 MP6 and the bpdbjobs output. I don't have the technote offhand, but it's been referenced on this list before. Try the list archives. The fix is a library patch from Symantec. -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] NBU & Aptare - Real World Benefits
* Matthew Johnson <[EMAIL PROTECTED]> [2007-03-20 13:57]: > I guess the overall consensus is that this is a good product. How much > does it cost It's not exactly cheap, but I think you have to talk to Aptare about the specifics. I get the impression that pricing is not necessarily a fixed option. :-( -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] NBU & Aptare - Real World Benefits
* Liddle, Stuart <[EMAIL PROTECTED]> [2007-03-20 11:39]: > Greg, > > I can tell you that we run the command: > > bppllist -allpolicies -U > > and put the resulting output into a file which we then put into a revision > control system (we use subversion). > > Plus we have a Change Control system where our users request changes to the > backups like adding/removing systems, path changes, etc. These work orders > are then referenced when we check in a change to the policy list. That is certainly a direct (and inexpensive) way to do it. Good idea. -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] NBU & Aptare - Real World Benefits
* Hindle, Greg <[EMAIL PROTECTED]> [2007-03-20 14:27]: > David, > How do you track policy changes? They have recently added a policy audit module that you can access the same way you would accecss an automated report. We are currently on 6.0.14, I don't remember off the top of my head when they added the access to the report. The reports are in the /opt/aptare/database/custom_reports directory in a typical install. -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] NBU & Aptare - Real World Benefits
* Ed Wilts <[EMAIL PROTECTED]> [2007-03-19 22:23]: > > These are benefits from StorageConsole that we get all the time - some > exclusively from StorageConsole and some in combination with our own > tools. I'm pretty sure that there are features that we're not using yet > (some just because they're not as applicable to our environment as they > could be to others and perhaps some just because I haven't mastered it > all (like automated reporting)). I can mirror most of this, plus add a few more. One of the more recent additions is tracking of policy changes. We are just starting to use this as an audit control of what changes are made in a given policy. It makes for a good balance check against our change management process. We have also saved many hours of our operations staff's time by giving them a single screen to monitor off-hours rather than have seven distinct NBU Activity Montiors running to track job status. The consolidated views simplify and help target what we react to. My favorite report is probably the Backup Duration SLA report. It gives a simple view of what backups are taking a long time to complete, allowing us to target those systems as potential items to optimize. -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] add_slave_on_clients
* Bob Stump <[EMAIL PROTECTED]> [2007-03-15 11:10]: > Hi Patrick, > If I remember right, when you and I were with OpenVision/VERITAS > ancient days ago, the add_slave_on_client running on the master could > be used to update all of the UNIX clients with a new media server > name. It wouldn't need any parameters. If anyone still has a NetBackup > 3.4 then they could verify the bourne shell script. Unfortunately it > is now an executable :-( > > Basically it used "bpgp" to collect all of the UNIX bp.conf files onto > the master and grepped for "SERVER = " and appended the new server as > "SERVER = NEWSERVERNAME". The it again used bpgp to push the bp.conf > back to the client. It always looked nasty because you then had > multiple entries for media servers in the new file. They should have > done a sort -u or something. If that's what it does, then using bpsetconfig is probably a better option. You have better control over what actually gets updated that way. -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] bperror -backstat -hoursago 24 does not report alljobs
* Sixbury, Dan <[EMAIL PROTECTED]> [2007-03-13 21:56]: > A couple of different issues... One is that cron very likely has to be > restarted to fix issues with time of cron jobs. In my case I had the > OS patched as well as the java in netbackup patched, however, I am > still seeing jobs report 1 hour off. I have also restarted the > netbackup daemons which didn't seem to fix the issue. Did you update the java on the client where you are running the GUI? -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] 58 Error
* Jimenez, Daniel <[EMAIL PROTECTED]> [2007-02-28 19:59]: > Hello everyone > > I have a question regarding an error I am receiving on a server running > netbackup 5.1 on a windows NetBackup environment . I'm attempting to > troubleshoot an issue with a server that keeps failing backups with a 58 > error. I was able to telnet to port 13724 from the client and was able > to telnet to port 13782 from the media server. When I open up the host > properties for the server on the master server it opens up fine. I also > created the bpcd log directory on the client and a log file was created > but not sure what to look for in the log file that will tell me if there > is a problem. I am new to this in depth troubleshooting so any help > would be appreciated, thanks in advance. Is the media server a system different from the master server? Make sure all systems are resolving DNS-type data correctly both forward ad reverse. You mentioned that you tested telnetting to port 13724 from the client to the media server, but did you verify that your client is configured to use vnetd? -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] 6.0 Java Console to manage 5.1?
* Paul Keating <[EMAIL PROTECTED]> [2007-02-28 11:29]: > LOL. > > Amazing to see a bunch of people so excited about a Unix desktop. > > ;o) You have no idea ;-) -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] VSP on windows Clients
* Staub, Doug <[EMAIL PROTECTED]> [2007-02-15 11:31]: > Also possible with the 'bpsetconfig' command as follows: > > echo "VSP_Use = NO" | bpsetconfig -h > then > bpgetconfig -M | grep VSP This doesn't work with newer (>5.1MP3?) anymore :-( -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] VSP on windows Clients
* Justin Piszcz <[EMAIL PROTECTED]> [2007-02-14 15:15]: > Yes, you click master server properties -> add the windows client to the > 'client settings part' and then you can turn vsp off in the second tab. The CLI version of this is bpclient. You will want to set the WOFB_enable to no (0). -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] Moving to a different storage unit
* Adeday <[EMAIL PROTECTED]> [2007-02-14 00:15]: > I have 4 disk storage units and a robotic library configured. One of the > disk storage units is full. I have 3 policies that are backing up to this > storage unit. What is the safest way of moving these policies to another > storage unit? Can I leave the old backups of these policies where they are > now and start backing up to another disk or tape? Will NBU know where the > old backups are? This is NBU 5.1 on solaris master/media server. Just change it in the policy. The only dangers you have are related to vaulting. If you aren't vaulting, then it's a non-issue. The original backups will stay related to where they originally backed up. -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] status 202, 205, 219 and 58
* Paul Keating <[EMAIL PROTECTED]> [2007-02-14 10:32]: > Yeah, that's definitely not the case. > > So since our core network is not going down, I have no reason to believe > it's a network issue. ;o) Our system is set up the same way, but we are seeing network-related errors during vaulting between the master server and the media servers. I doubt our 6509 is the cause for this, either ;-) -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] NBU and Aptare custom reports ..
* Philip McDougal <[EMAIL PROTECTED]> [2007-02-13 12:03]: > Thanks Paul. I've tried that already ;-) but they want a > commitment from my company before they 'allocate resources' to do > this. So I'm almost forced to do this myself. > > I love the product and want to have the company purchase it but > there's no way they're gonna like me telling them that the most > important report we need is one that WE have to create. You _can_ work out with them that they need to supply a particular report as part of your agreement within a given period of time. That worked out fairly well for us. -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] NetBackup and New Daylight Saving Time Rules
* Ed Wilts <[EMAIL PROTECTED]> [2007-01-10 18:12]: > On 1/10/2007 11:11 AM, Jack Forester, Jr. wrote: > > This isn't something I've seen discussed here in recent memory, but it's > > something we're working on right now. The new Daylight Saving Time > > rules for the US take effect this year and we're wondering if anybody > > has investigated the impact to NetBackup. > > I've asked Graeme at Aptare to comment as well since StorageConsole is > java-based. When I hear back, I'll post to the list (or one of you > Aptare guys can finally learn to post :-)). Somehow, I don't think that's ever going to happen ;-) -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] solaris client version level
* Curtis Preston <[EMAIL PROTECTED]> [2007-01-04 12:58]: > You can also do this without ssh: > > for box in `cat customer-box-list` > do > echo "Machine [$box] level is:" > bpgetconfig -M $box |grep -i ver > done Yes, but that is horribly inconsistient and nowhere near being reliable. At best, you can usually get the base level of the version that is installed, but you won't get any information about any maintenance packs. I have found that bpcoverage does a better job, but it's still not perfect. -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] solaris client version level
* Ed Wilts <[EMAIL PROTECTED]> [2006-12-30 08:50]: > On 12/30/2006 12:00 AM, David Rock wrote: > > * Bobby R Windle <[EMAIL PROTECTED]> [2006-12-29 22:43]: > >> Does anyone know how we can confirm what maint pack level we are running > >> on our Solaris clients. I tried to find this out > >> before but never got any hits. Is there a version command or something we > >> can run. > >> > >> Running Netbackup Enterprise 5.1 mp6 on media servers. Backing up all > >> platforms. Need to find what version our Solaris clients are on. > > > > NetBackup is notoriously bad for figuring out what version you are > > running. A couple things you can try are bpcoverage, bpgetconfig or the > > NCVU http://seer.support.veritas.com/docs/276337.htm > > With a Solaris (or other Unixy) master server, why fight the battle? > Simply push out new clients out to _all_ of your Solaris clients. > NetBackup will skip over the clients that are up to date and update > those that need it. The update to a Unixy client is trivial and usually > transparent although I am waiting for the 6.5 release where Symantec has > done a MUCH nicer job of handling client updates. That sounds like a great idea.. oh, wait. I can't use rsh because that's blocked by the firewall and it's turned off on most of the clients anyway because it's insecure on many levels (so is ftp). I can hack the script to use scp instead, but why should I have to fix something that should have been done a loong time ago by Veritas/Symantec? Sorry, spinal reflex/complaint. There are so many. ;-) I think it's pretty obvious I'm not afraid to hack stupid things to get them to work right. Unfortunately, you can also have issues when you are not the admins of the client systems. And then there's Windows... Realistically, the biggest reason for NOT being able to do that is change control. Just asking the client what it has does not involve two days of justifcation and meetings about why you want to update clients. :-( So, one more plea for someone at Symantec to get off their behinds and actually UPDATE the [EMAIL PROTECTED]@! version strings, please!! I'm going to stop beating the dead horse, now. Happy New Year, everyone (yes, even the folks at Symantec). I do realize trying to make something as big as NBU work is a pain. :-) -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] solaris client version level
* Bobby R Windle <[EMAIL PROTECTED]> [2006-12-29 22:43]: > Does anyone know how we can confirm what maint pack level we are running > on our Solaris clients. I tried to find this out > before but never got any hits. Is there a version command or something we > can run. > > Running Netbackup Enterprise 5.1 mp6 on media servers. Backing up all > platforms. Need to find what version our Solaris clients are on. NetBackup is notoriously bad for figuring out what version you are running. A couple things you can try are bpcoverage, bpgetconfig or the NCVU http://seer.support.veritas.com/docs/276337.htm -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] Firewall setup
* Anas Kayal <[EMAIL PROTECTED]> [2006-12-21 09:29]: > Guys, I have 2 servers in my DMZ. Now after reading this forum I opened > ports 13720 and 13724 and permitted access from my master server to both > servers in DMZ in both directions. Now how do I specify that this port > should be used by this client and the other port by the other client? Don't forget that you also need 13782 (bpcd) from the backup server TO the client. 13720 (bprd) and 13724 (vnetd) are for traffic FROM the client TO the backup server. -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] NetBackup Buffers
* mkiles <[EMAIL PROTECTED]> [2006-12-05 10:35]: > Hi All > > Has any one done any benchmarking with > NUMBER_DATA_BUFFERS, SIZE_DATA_BUFFERS, > NUMBER_DATA_BUFFERS_DISK, SIZE_DATA_BUFFERS_DISK, > NUMBER_DATA_BUFFERS_RESTORE, SIZE_DATA_BUFFERS_RESTORE > and NET_BUFFER_SIZE with NBU 6.0 > > I read it somewhere that Veritas has upped the default > values for these tuning variables and creating those > entries in /usr/openv/netbackup/db/config is no longer > neccessary. haven't worked with 6.0 yet, but you could get at least some of the basic information by looking in the logs for information on what value is being used (e.g. bptm, bpdm, etc). -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] RES: Master server merging
* Carlos Alberto Lima dos santos <[EMAIL PROTECTED]> [2006-12-01 11:22]: > With NBU 6.0 I didn't try, but until the version 5.1 I know that is > possible, you need only be careful with clients with the same name in both > masters and change the configuration in all clients to point to new mergerd > master. Sure, but like Ed said, there is NO _supported_ way of doing it. It might work, probably will work, but you are on your own if it doesn't. -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] Aptare gotchas?
* Tschida, Tom (STP) <[EMAIL PROTECTED]> [2006-11-15 12:05]: > Hello all, > > we're looking at purchasing Aptare for our NetBackup and BackupExec > (yes, they will be supporting BE this starting this month) environments. > It's it promoted for the most part as easy plug-n-play. Does anyone > have any lessons learned regarding the implementation of this tool? Any > info is appreciated. The implementation is generally pretty easy, we have done installs and upgrades over the last few years with only a few small issues. The biggest lesson learned I can give you is to read the documentation throughly BEFORE starting. It is pretty well laid out and not too hard to follow, but there are small things that you could miss if you don't pay attention. None of them are earth-shattering, but can be a little annoying. :-) Their phone support for the install is very good and the whole process took only a couple hours for us. -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] Bocada Reports - Please Help!
* Ellis, Jason <[EMAIL PROTECTED]> [2006-10-18 09:29]: > A while ago I posted up about Aptare versus Bocada. The story behind the > e-mail was that we already had Bocada, however our new manager was > looking to replace it with Aptare. We have a demo install of Aptare and > I'm very impressed with it, unfortunately it looks like we will be stuck > with Bocada for the next year and I've been tasked with working with it. > > If anybody happens to have any good Bocada reports they've created for > NetBackup (SQL queries I believe) that you're willing to share with me, > please contact me off list. Thanks! I guess the silly question is, have you tried asking Bocada? Maybe what you need to get out of it is already there and you just don't know where to look. Or at the very least they can point you in the right direction without charging for professional services. -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] 9940B
* Clooney, David <[EMAIL PROTECTED]> [2006-10-14 15:19]: > Keep your hair on Bob > > Ok so tyou know about drives, point taken. > > I think you misinterpretted my mail. > > I was merely trying to find out the maximum I/O the drive supports , > phew. We have seen real world testing at 55MB/sec and I believe the rated speed is around 65MB/sec -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] Trying to implement synthetic backups
* Carlisle, D Renee <[EMAIL PROTECTED]> [2006-10-10 13:26]: > A "Synthetic Full" uses the same amount of space as a real full. You > can save a little bit of space during the week if you use differential > incremental versus cumulative incremental, but that would depend on > your change rate. I personally, would not use synthetic backups in a > tape environment. If you do your incremental backups to disk and only > the full's to tape, that would be ok. If you try and create a > synthetic full off of a weeks worth of tapes it will be cumbersome and > slow. > > The biggest advantage for us with Synthetic full backups is that our > backups that used to run for 12 hours on our client systems are now > off loaded to run on the media server as the Synthetic is being > created. Yeah, the benefit is saving of _time_, not tape. When the Synthetic backup is created, it's all done on the media server. The client is no longer involved and the "new" full will be created as fast as your media server can go, but it _will_ use a full backup's worth of tape. -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] Has anybody used Aptare?
* Dave Brown <[EMAIL PROTECTED]> [2006-10-10 13:27]: > > Thanks for the info, I will call and get the demo One thing I will add. Storage Console (Aptare) does have a GB/Month detail, which also includes GB duplicated. We are currently working through with them to figure out exactly what that means. We have recently come across a difference of opinion about what the report _does_ show .vs what it _should_ show. It _does_ accurately show the GB backed up in a month, and all GB duped for the same client in that month, but it's only as good as the data NetBackup supplies. The main issue is with Virtual tape being used for staging. If you do NOT want that data to be counted, that has not been completely ironed out yet. Again, not a bad thing, you just need to have a good feel for what questions you _really_ want to answer. You will get a lot farther if you can do that up front. Vague questions are hard to respond to. Incidentally, you said you were using Bocada. Why are you now looking at Aptare? Is there something you need that it can't do? -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] NetBackup Schedules - 3 Schedules in 1 Policy advice
* WEAVER, Simon <[EMAIL PROTECTED]> [2006-10-06 06:35]: > > All, > > NBU 5.1 MP2 Win2k Master + 2 SAN MS > Morning! I am trying to work out a solution to trimming down all the > policies I have. Presently, I have the following:- > > 1 x Policy for Month backups, which contains 1 Schedule. Schedule starts 7pm > Fri and ends 7am Sat > > 1 x Policy for Weekly Backups, which contains 2 Schedules. 1 x Schedule is > the Full Backup, starting 7pm Fri and ends 7am Sat. > The next schedule is for Incr which take place 7pm - 10pm Mon, Tues, Wed and > Thursday only. > > If I was to place the MONTHLY SCHEDULE (Copy / Paste) into the Weekly Backup > Policy, would I get the Monthly and Weekly full backups trying to run at the > same time? > > I am trying to work out the best way (and effecient) to ensure monthly runs > on a set day only, but does not interfere with the regular weekly full > backup. Adding the Monthly schedule to the existing policy containing the Weekly and Daily schedules will not result in both running at the same time. Frequency based scheduling does just that, it give priority to the schedule that runs on the _least_ frequent schedule. So, if there is a conflict between the Monthly and the Weekly, the Monthly will win. Having them in separate policies _guarantees_ they will _always_ try to run at the same time. Having them in the _same_ policy is specifically intended to manage it for you so they don't. Of course, you will need to play with it to make sure it does what you think it's supposed to do :-) -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] Using Different Hardware Platforms for NetBackup MediaServers...
* WEAVER, Simon <[EMAIL PROTECTED]> [2006-10-05 07:05]: > > Yes - take your pick :-) Not only yes, but often required. Try to mount a Windows-based BCV snapshot to a Solaris media server for backup sometime :-) Seriously, one of the main reasons for having different hardware/OS media servers is for specialized OS/filesystem requirements, like splitting a mirror on an HP system and mounting it to the media server for performance. You will usually need to have the media server be the same type of OS as the client in order for that to work. -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] bpclient
* Tairone N. Magalhaes <[EMAIL PROTECTED]> [2006-10-04 18:32]: > Thank you Stuart. It seems that bpclients lists only the clients that were > previously added using the same command. The command bpplclients is exactly > what I was looking for, it lists also the OS of the clients, and I need it > to configure the sftp only on UNIX clients. > Thanks, bpclient is for changing client properties on the Master server (setting allow non reserved ports, vnetd, etc.) -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] Client Hostnames
* Hillman, Eric <[EMAIL PROTECTED]> [2006-09-29 12:44]: > Thanks for all of your replies. This has been a very interesting and > helpful thread. I may have found one issues when using aliases however. > Can anyone confirm that you CANNOT use an alias of the server name when > backing up an Exchange server/cluster? From what I'm reading, the > netbackup client needs to reference the Exchange application using the > actual hostname used when installing Exchange in order to successfully > access the specified storage groups. This would mean that the actual > hostname of the box needs to be added to the Netbackup Policy and not a > "-backup" hostname. Hopefully, this would be the only case that an > alias cannot be used. Thanks! I can confirm that Exchange gets very unhappy when the client name does not match the actual hostname. -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] Client Hostnames
* Bobby Williams <[EMAIL PROTECTED]> [2006-09-28 05:47]: > > About the FQN's, we don't use them. Why type all of those extra characters? > Our DR strategy is to carry our offsite tapes to one of our other sites and > do our recovery. Without using FQN's, we just bring a development box back > up as a lost production box and put it in the domain of the DR site. Depends on environment. We have a shared NBU infrastructure that backs up multiple client environments, therefore we have the potential problem of two different clients with the same short name. -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] help with netbackup
* "Koster, Phil" <[EMAIL PROTECTED]> [2006-09-28 09:51]: > > This really would not be an issue if it was a new system or presented a > little differently. But when you inherit the system and some of those > lists are partially (or nearly fully) populated from a predecessor, you > can get confused. Particularly if the last person set the system up > wrong and over complicated your set up. NetBackup with an overcomplicated set up??! Never!!! :-) Yeah, NBUs choice of what to automatically populate or not is insane. One thing to remember along those lines that will drive you nuts is almost all the client information is NOT polled from anywhere. The Hardware and OS entries are simply picked by the admin setting up a policy. If the client system changes OS, you'll never know. Take almost everything NBU tells you about your clients with a grain of salt. :-( -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] Client Hostnames
* Hillman, Eric <[EMAIL PROTECTED]> [2006-09-27 14:37]: > We have public and backup networks. All clients have a public network > which is typically 100Mb and some clients (larger clients) have a backup > network which is typically GbE. This seems to be pretty typical in > larger organizations. > > Given the above scenarios, my question is... in Netbackup what do you > name your clients with backup nic's? We name the client the same no matter what and make sure the routing on the client is correct. Where things start to get weird is when you are using production-side dns to resolve backup server names can cause split identity problems. -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] how to disable manual backups during the daytime
* Bobby Williams <[EMAIL PROTECTED]> [2006-09-26 12:17]: > Why don't you go into your schedules that are UBAC (user backups) and > have the start window only open for after hours? > > You can search to see what policies have schedules of type UBAC by > using bpplsched. The issue isn't User-Initiated, it's immediate jobs started by an admin. Setting global retries to 0 won't do it, we use that for DR to keep scheduled things from running, but manual stuff still goes. Deactivating the policies would work, but it's dangerous. What happens if the cron job fails to re-activate the policies? Not sure what else would work off the top. :-( -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] Netbackup 5 to 6: deal or no deal
* Siano, James C <[EMAIL PROTECTED]> [2006-09-23 13:39]: > Netbackup 5.0 mp6 to 6.x mpX > > Are there really outstanding, compelling features or reasons to move > from NBU 5 to 6? Unix master backing up unix/wintel variants. If you aren't aware of any, then no. :-) Seriously, If things are working for you now, then there is no reason to change. Upgrading to a new version just for the sake of doing it is generally a bad idea. Staying current on patching for the version you are using is not. You have to ask yourself "do I have any outstanding issues that are addressed in NBU6?" If the answer is no, then wait. -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] How do we duplicate skipped images in vaulting
* Veritas Netbackup <[EMAIL PROTECTED]> [2006-09-23 01:41]: > Hi All, > > We have partially completed Vaults (error 1), its possible that the VTL > media is in use when the Vault is attempted or that the image expired before > the vault could duplicate it onto the Tape. > > How do I find a list of such images, so that I can refire the backups and > reduce the risk. Incase of a monthly backup this is more important. > > We use the foll vaults > 1. Weekly vault 3. Quarterly Vault > 2. Monthly vault 4. Offsite Vault. > > Currenlty we are not sending all backups offsite. If you set up the vault profile correctly, it will try over several days to vault the images that have not successfully vaulted. Is that what you need it to do? -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] Exclusion lists - How to list all on server
* Bob Stump <[EMAIL PROTECTED]> [2006-09-21 07:55]: > The sticky part is where there are more than one exclude_list > exclude_list.Policy > exclude_list.Policy2.Sched1 > exclude_list.Policy2.Sched2 Yep, you will just need to loop through those and pull them back to look them over. -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] Exclusion lists - How to list all on server
* "Koster, Phil" <[EMAIL PROTECTED]> [2006-09-20 16:21]: > For those keeping track, I am trying to audit our back up policies (NBU > 6 MP2 on Win 2K Srvr). > > I have almost everything except I can not find out how to list all the > excluded directories or files. I would like them all in one nice neat > clean list or at least somehow CLI so I can parse it. > > Any thoughts other than I am stuck doing this one server at a time > through the GUI? This one gets beat to death on the list. :-) Basically, yes and no. For Windows clients, you can use bpgetconfig to retrieve Exclude data. ( bpgetconfig -M client ). For UNIX clients, it's a little more complicated. You need to be able to collect all the exclude_list.* files in /usr/openv/netbackup to see what is excluded. You can try using bpdir to verify _if_ there are exclude files and then use scp or some other method to get the contents. -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] Bppllist columns
* Clooney, David <[EMAIL PROTECTED]> [2006-09-19 11:53]: > Hi All > > Does anyone have the columns for > > bppllist $policy > > Basically trying to put something together based on policy type, trying to > get the columns specifically for the INFO section below. > > Eg. > > CLASS croyvtsms_oracle *NULL* 0 50 169200 *NULL* > NAMES > INFO 0 0 0 0 *NULL* 0 0 2147483647 1 0 1 0 0 0 0 0 0 0 1041511308 > 3E0E2989A5AF4185B6575FE1DA53C67A 1 0 0 0 0 0 0 0 2 This is what I have gleaned from working with it. I believe it covers most of what you probably need to know. I'm still working on the rest :-) # Policy INFO line # 1 2 3 4 5 67 8 910 11 12 13 14 15 16 17 18 19 20 21 22 #INFO 6 0 0 0 *NULL* 0 0 2147483647 0 0 0 0 0 0 0 0 0 0 1088520139 85BD1B0A1DD211B2AE2C0800208094C0 1 # # 1 Tag # 2 Policy Type # 3 Follow NFS/Backup Network Drives (0 no/1 yes) # 4 Compression (0 no/1 yes) # 5 Job Priority (int value) # 6 # 7 # 8 Collect DR Info (0 no/1 yes) # 9 Limit Jobs Per Policy (int value) #10 Cross Mount Points (0 no/ 1 yes) #11 #12 Active (0 yes / 1 no) #13 Collect TIR (0 no/1 yes/2 yes with move detection) #14 #15 #16 #17 Allow Multiple Data Streams #18 #19 #20 Active Date #21 #22 -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] Policy Auditing: How to
* "Koster, Phil" <[EMAIL PROTECTED]> [2006-09-18 09:14]:
> NBU 6 MP2 on Win2K Srvr
>
> Me and the other "backup admin" inherited the NBU system from someone
> else. We want to audit the policies. I think I already know the answer
> but is there any "easy" way to get a list of which servers are in which
> back up policies without having to do this one policy or one client at a
> time? It'd be nice to get it dumped to like a csv or something we can
> process to look for duplicates etc.
bppllist will give you a list of policies and their configs. this
includes the clients, schedules, etc. Use the "-U" for a more human
readable form.
bppllist: [] [-L|-l|-U] [-verbose] [-allpolicies]
[-M ]
[-hwos]
[-byclient ] [-keyword "keyword phrase"]
--
David Rock
[EMAIL PROTECTED]
___
Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu
http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] Backup through firewalls
* Whelan, Patrick <[EMAIL PROTECTED]> [2006-09-15 15:34]:
> I have a script that runs bprdreq -rereadconfig 50 times (50 is
> completely arbitrary) and so far it has worked every time.
I have never had a problem with on-the-fly changes for vnetd. You can
also use the CLI to make these changes a LOT faster than screwing with
the Java GUI.
Specifically, the -no_callback option in bpclient
USAGE: bpclient -All [-M master_server] [-l|-L|-H|-FI]
bpclient -client client_name [-M master_server]
-l|-L|-H|-FI
bpclient -client client_name [-M master_server]
-add|-delete|-update
For -add and -update the options are
-connect_nr_port <0=no, 1=yes>
-no_callback <0=no, 1=yes>
-dynamic_address <0=no, 1=yes>
-free_browse <0=allow, 1=deny, 2=use>
-list_restore <0=not specified, 1=allow both,
2=allow list only, 3=deny both>
-max_jobs <1-99>
-current_hostname
-current_ip_addr
-current_host host_name[:ip_address]|:ip_address
-WOFB_enabled <0=WOFB disabled, 1=WOFB_enabled>
-WOFB_FIM <0=VSP, 1=VSS>
-WOFB_usage <0=Individual Drive Snapshot, 1=Global Drive Snapshot>
-WOFB_error <0=Abort Backup on Error, 1=Disable & Continue>
--
David Rock
[EMAIL PROTECTED]
___
Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu
http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] Backup through firewalls
* smpt <[EMAIL PROTECTED]> [2006-09-15 07:05]: > Hi, > I've configured some firewaled NetBackup domains with vnetd and I never had > any problem with streams. > > I have ages to hear from someone the port model. I had proposed this to some > of my customers and when the firewall admin understood how many ports needed > they refused it immediately. Yep. The only reason we had them at all was because of legacy firewall configs for NBU 3.2 and 3.4. We have been trying to get rid of all port range stupidity for several years, but it's always the old "if it ain't broke..." Incidentally, ACSLS 7 can be configured for single port communication over TCP only, too. That was another big pain. -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] Command for finding out witch files are notbe ingback-up with status code 1
* Clooney, David <[EMAIL PROTECTED]> [2006-09-15 13:09]: > What you need to do is > > bperror -problems -client $client -d mm/dd/ (start time ) -e > mm/dd/ (end time) -U > > Start and end time being either side of when the backup took place . > > If there was an error code 1 (partial sucess) then netbackup will have a > record of it is is still in the error DB Now that I think about it, this information is also in the bpdbjobs -all_columns output. It's a little more complicated to get it out, though. It's buried in the try status data information. -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] Backup through firewalls
* [EMAIL PROTECTED] <[EMAIL PROTECTED]> [2006-09-14 13:48]: > There's a whole section on this in the SAG. > > Shortanswer, you need "bpcd" from the master or media server to the > client, "vnetd" the reverse direction. You have to make sure you > configure the client for "no callback connections" via the bpclient > command or, no doubt, someplace in the GUI. > > Users on the client cannot perform their own restores using this. I'm > told, but have not verified, that you can enable "bprd" from client to > master to allow this. Speaking as a backup guy who is now on the firewall team, using vnetd is by far the recommended way of dealing with the firewall. If all you are dealing with is backup servers to client machine, the short list is: Server -> Client port 13782 (bpcd) Client -> Server ports 13724 (vnetd) and 13720 (bprd) Yes client initiated restores will work with just these ports. If your backup servers are hanging off of a DMZ so that your admin clients using the Java GUI need to get access, you can also use: Admin Client -> Server ports 13722 (bpjava) and 13724 (vnetd) This will also require the /usr/openv/java/nbj.conf file setting of NBJAVA_CONNECT_OPTION=1 (default is 0) The only downside to vnetd that I have heard of but not seen personally is that you are limited to a single stream for backups, which could impact your backup model if you are trying to use NEW_STREAM file directives. If that is the case, you can configure port ranges and I highly recommend using ALLOW_NON_RESERVED_PORTS as part of that. Using low ports (<1024) by default is one of the stupidest things NBU ever did. -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] Command for finding out witch files are not being back-up with status code 1
* Edwin Bader <[EMAIL PROTECTED]> [2006-09-14 16:35]: > Gents, > > > Several customers are asking witch files are not being backed up when > the status code is 1. > > Is there a way to retrieve this info? bperror -client -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] backup 1K files that can't be restored? (was Re: Fame at last.)
* Rob Worman <[EMAIL PROTECTED]> [2006-09-08 17:28]: > There is one sentence that confuses me about Ms. Fisher's > recent CW article: > > "One major issue Veritas users cited with NetBackup 6.0 is > that when a backup job indicates it's complete, it often has > only backed up 1K files, so there's no data to restore. " > > Whoa, that DOES sound scary. Backed up files that can't > be restored?!? > > But what does that sentence really mean? Especially the > reference to "only backed up 1K files"? Is that meant to > symbolize "1000 files"? > Another example of a similar symptom is a backup of a mountpoint where a filesystem is supposed to be mounted, but never did so the resulting backup is only of the directory mountpoint itself. Status 0, but no files were backed up. That also would not be NetBackup's fault. It backed up exactly what you told it to, there just wasn't anything there when it tried. Note that is different from a Status 71 where specific files were listed and were not found. Both of these are common errors when doing user-directed database export backups or bcv mounts. The script to create the export or mount the bcv fails, runs the backup on an empty directory and poof, a successful failed backup. User error or bad script design, not a NetBackup bug. Some of the current reporting packages are capable of flagging such a situation to alert you that _maybe_ something is wrong. -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] any suggestion about this
* Kohli, Vidit <[EMAIL PROTECTED]> [2006-09-08 13:35]: > what this means on master / bpps -a, > its SSO , robotic controlled by master > > MM processes > > root 16080 1073 0 13:30:21 ?0:00 > /usr/openv/volmgr/bin/oprd -sockfd 7 > root 15907 1073 0 13:28:42 ?0:00 > /usr/openv/volmgr/bin/oprd -sockfd 7 > root 187 1073 0 11:13:16 ?0:00 rdevmi -sockfd 7 -r > root 15857 1073 0 11:24:51 ?0:00 rdevmi -sockfd 7 -r that looks like vmd is not running somewhere or a media server is not up. -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] Policy Export / Import?
* WEAVER, Simon <[EMAIL PROTECTED]> [2006-09-08 06:42]: > > David > That sound like A LOT of work for what seems to be a simple process?? > Does a copy and paste work Jonathan ? I suggest it because it would probably be the only "supported" way of doing it. There are several things that _should_ work, _might_ work, etc. that will definitely NOT be supported by Symantec. It's not quite as bad as it sounds, but it does work and does involve some time. The only time it would really be worth it is in an environment where you may need to do it more than once, or for the altruistic sake of "because it's there". If it will need to be repeated, it should probably be scripted. :-) To be fair, I _have_ done a lot of scripting work in the past for this kind of stuff, so the perception of what constitutes a LOT may be slightly different for me ;-) -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] Fame at last.
* Ed Wilts <[EMAIL PROTECTED]> [2006-09-07 18:49]: > > Some NetBackup engineers do read this list (or get relevant messages > forwarded to them). So do some of the support folks (ditto). From what > I've seen and personally experienced, they're trying really, really hard > to fix their early shipping screwup. I personally have absolute no > issues with their support organization nor their efforts to make things > better. I have to agree with Ed. The engineers are just people doing a job like the rest of us. I'm sure no-one here on the list has _ever_ had a complaint from a customer for their own work :-) The individuals are not the problem, the behemoth that NBU has become is. I have personally seen that there are people that care about their product, but it _is_ a bear to fix some of the issues. That doesn't make it any easier for us, but I don't think they are out to deliberately screw us. -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] Policy Export / Import?
NetWare MS-Windows-NT OS/2 MS-Exchange-Server DB2 AFS Lotus-Notes FlashBackup-Windows Vault -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] Fame at last.
* Mansell, Richard <[EMAIL PROTECTED]> [2006-09-07 12:39]: > I see some of you were also contacted by Sharon at Computerworld:- > > http://www.networkworld.com/news/2006/090506-users-suffer-major-problems > -with.html There was also an article in Storage. -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] User Restores?
* Martin, Jonathan (Contractor) <[EMAIL PROTECTED]> [2006-09-06 16:35]: > > Is there any feature of NBU that would allow users to restore their own > data? I suppose we could give them access to the restore GUI, but they > could then restore any file to any location, which is a major security > issue. (Whoops, did I just overwrite the kernel?) Is there some sort > of way we could give users the ability to restore their own data without > the DLO option? If you don't trust them to restore their own data, then don't let them do it. You can't really have it both ways. I am assuming they are admins of the boxes they are restoring files for. Odds are much higher that they will shoot themselves in the foot some other way than screwing up a restore :-) -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] Migration from 4.5 FP3 to 6.0
* Yang Xiao <[EMAIL PROTECTED]> [2006-09-06 14:02]: > Hi all, > After talking to Veritas support, I guess since we are running 4.5 on RH > 7.3, > and non of them is really supported, they suggested that we just build the > new server with ES 4 and NB 6.0 MP3, and import all of the tapes to rebuild > the catalog, since they said the 6.0 should read the tapes from 4.5. > Any thoughts on this? "ouch?" How many tapes are you dealing with? If it's a relatively small number, you could get away with that. Make sure you have a dump of your images and their retentions to help with researching images to import in an emergency and then work on a schedule to get as many as you deem important imported as quickly as possible. You may want to also try and identify critical systems and import those tapes first. -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] NBU 5.1 and best VMWare Host backup practices
* Scott Jacobson <[EMAIL PROTECTED]> [2006-08-31 09:18]: > VMWare guru's > > For now we are going with direct Native or Standard Client > based backups of a Hosts "Guests" (working just fine actually) rather > that snapshots. > After scouring both WEB sites, I can only find information about > "snapshot" based backups for "Guests". > What I'm looking for are references to the "Hosts" backup, > specifically, what files need to be backed > up and used for a restore if I needed to lay down a new OS and then > restore the files > necessary that would allow me to then restore those Guests previously > defined on that Host. It's pretty straightforward fo the most part. Essentially you need to back up the config file for the vm and the vmdks. Some details that are important are what version of VMware, planning a schedule to put the guest in a quite mode so the disk files can be backed up, etc. You are basically doing a "normal" backup of the host system. Information about specific files will be somewhat dependant on which VMware product you are using, but the methodology is pretty much the same. -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] vnetd question
* Dave Lowenstein <[EMAIL PROTECTED]> [2006-08-24 15:58]: > I'm trying to make sure that netbackup 5.1 will work through a firewall > with minimal ports being opened. > > What ports need to be open on the firewall if filtering by destination port: > > Media >> Client > 13782 (bpcd) > > Client >> Media > 13724 (vnetd) > > If the client needs to run user backups/restores, then the following > port will also need to be opened: > Client >> Master > 13720 (bprd) > > > Right? So I'm testing this out between two hosts with some more liberal > firewall rules than above. Backups are working fine, restores are > working fine. > > I believe I have all the vnetd stuff set correctly, although every piece > of documentation I find shows a slightly different gui interface than > what I'm seeing. > > Why am I still seeing it talking back and forth between media server and > client with randomly selected destination ports (like 852 and 811)? > There's also a fair amount of icmp going on back and forth between the > two, which I'd like to be able to close down. > > > > client -> serverTCP D=13724 S=852 Ack=392120625 Seq=931198138 > Len=0 Win=49680 > client -> serverTCP D=13724 S=852 Push Ack=392120625 > Seq=931198138 Len=2 Win=49680 >server -> client TCP D=852 S=13724 Ack=931198140 > Seq=392120625 Len=0 Win=33118 >server -> client TCP D=852 S=13724 Push Ack=931198140 > Seq=392120625 Len=2 Win=33120 > client -> serverTCP D=13724 S=852 Ack=392120627 > Seq=931198140 Len=0 Win=49680 > client -> serverTCP D=13724 S=852 Push Ack=392120627 > Seq=931198140 Len=2 Win=49680 >server -> client TCP D=852 S=13724 Ack=931198142 > Seq=392120627 Len=0 Win=33120 > client -> serverTCP D=13724 S=852 Push Ack=392120627 > Seq=931198142 Len=21 Win=49680 >server -> client TCP D=852 S=13724 Push Ack=931198163 > Seq=392120627 Len=33 Win=33120 > client -> serverTCP D=13724 S=852 Ack=392120660 > Seq=931198163 Len=0 Win=49680 > client -> serverTCP D=811 S=13782 Push Ack=1618165286 > Seq=912667687 Len=2 Win=49680 >server -> client TCP D=13782 S=811 Ack=912667689 > Seq=1618165286 Len=0 Win=34500 What this looks like to me is that your SOURCE port is 852. Any real firewall won't care about this. What you are concerned about is if the DESTINATION is correct. What is most likely happening is the connection from the client to the media server has been established via vnetd, but the client end port is 852 or 811. The client source port doesn't have anything to do with it. The fact that the client is going after 13724 means that it's set up correctly. -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] VTL images to tape
* Martin, Jonathan (Contractor) <[EMAIL PROTECTED]> [2006-08-23 09:13]: > > Symantec Netbackup Vault is a software package that assists NBU Admins > in duplicating images from "disk to tape" or "tape to tape" for the > purposes of taking them offsite! While it isn't REQUIRED to do > duplication it does "assist" in making this job easier, especially if > you frequently send media offsite for storage. This is an "add on" > package for Netbackup and has been around since at least 5.0 if not > going back to 4.5 or farther. This one has been around since the 3.4 days as an extra. It existed even eariler as a professional services offering. It was bolted on in 4.5 and has existed ever since (with varying degrees of success). It sounds like the problem is not with Vault, but with the Exchange-related Enterprise Vault. Is that true? -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] VTL images to tape
* Dyck, Jonathan <[EMAIL PROTECTED]> [2006-08-22 14:04]: > Sounds like Greg and the some of the rest of the the mailing list are > likely talking about two different products here... > > Symantec Enterprise Vault now has a feature to migrate data to disk/tape > etc that was backed up with Netbackup 5.1+ > > "Vault", the add-on to Netbackup can dupe from VLT to tape... That's entirely possible. "Enterprise Vault" is for vaulting MS Exchange data, while "Vault" is the NetBackup image duplication. I am talking about "Vault". -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] VTL images to tape
* Liddle, Stuart <[EMAIL PROTECTED]> [2006-08-22 10:25]: > Actually, the story I heard from Symantec/Veritas about the VTL support > under 6.x is that it will allow for automatic expiration of the VTL images > once they have been successfully duplicated to physical tape. That's > similar to the current option in vault for images coming from a DSSU. Yes, that's one of our enhancement requests (like many others) because we use the VTL mostly for staging only and would really like a checkbox in vault like the one for DSUs to expire after X hours. -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] VTL images to tape
* Hindle, Greg <[EMAIL PROTECTED]> [2006-08-22 12:55]: > Yes using as tape. All that works fine. But getting the images from the > VTL to tape is more challenging. I think netbackup 6.5 has full support > for VTL's and then I assume that the vaulting option will fully support > VTL to tape copying then. I was told that 6.5 is due out by the end of > this year. That sounds like complete garbage to me. We have been using VTL (Quantum DX100) since 4.5 and have had no issues at all with vaulting to "real" tape. Quite the opposite, it has helped our environment tremendously. We are on 5.1MP[45] for most of our stuff now. What, exactly, are you being told by Symantec? There _are_ issues with how we are trying to manage getting stuff expired, but that has nothing to do with bpduplicate. -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] Quantum DX100
* Hall, Christian N. <[EMAIL PROTECTED]> [2006-08-16 08:57]: > Wow. Inline hardware compression is one of the reasons we purchased it. > Why have you been not able to get the compression cards to work? We came into it very early, for one thing. I think we were one of the first companies to use it, if not _the_ first. There were several bugs in the hardware that took quite a while for them to address. I don't know off hand if we ever got a chance to go back to trying it again. The issue was that you have to reformat the box to make the switch, so if it doesn't work, you need to bpduplicate ALL the images off to real tape before the compression cards can be swapped for normal cards or vice-versa. I have been out of the backup team officially for almost a year, so I can't speak directly to current hardware status. Maybe they've fixed it? I would keep it in mind and try testing heavily if possible. You may want to also mention it to Quantum to see what shakes loose ;-) -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] Quantum DX100
* Hall, Christian N. <[EMAIL PROTECTED]> [2006-08-15 17:11]: > All, > > We have purchased a DX100 for our Netbackup environment is there anyone > using this product and what has been your experience with it? We have been using it for a couple of years now and have been very happy with the hardware and the overall performance. The main thing we use it for is staging before writing offsite tapes. We have never had good luck with the compression cards, though. -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] bpgp ?
* Clooney, David <[EMAIL PROTECTED]> [2006-08-15 12:35]: > Hi all > > Trying to rollout an upgrade of x number of solaris clients , the > problem I have is the SA'a want to simply remove the package and then > add the new, which will detroy any exclude_lists > out there in the environment. > > I use bpgp quite a bit in secured environments and find very useful > indeed, albeit I have seen technotes in the past it cause issues I have > yet to have a prob in the last 4 years. > > Problem is that you have to know the specific file you are after on the > client, has anyone incorpated wildcards in some sort of way or fashion > to bring down say, > exclude* ?? Ack! bpgp has no way to use wildcards. If you do try to use wildcards, you will likely destroy something. Always be very very careful when using bpgp. -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] Backward Compatible
* Clooney, David <[EMAIL PROTECTED]> [2006-08-15 15:05]: > Firstly, i would just like mention that the expertise within this group > are much appreciated > > This subject I suppose has sprung up from the bpgp forum that has > bouncing around and local SA's realising they have to upgrade. > > Question is > > Master server is being upgraded to NBU 6.0 MP$ in the near future and it > has become apparent that there are a load of NBU solaris 2.6 clients > running 4.5 mp$ . solaris 2.6 is no longer supported as of 5.0< , so > would it be possible to to keep a 5.1 mp$ media server backing up > various solaris 2.6 NBU 4.5 mp$ ??? which still hooks back in to the NBU > 6 master server. > > I know this is a dubious question, but the ref "not supported", backups > would still be completed wouldn't they ? That should actually be supported. Our 5.1 Master -> 4.5 Media -> 3.4 client was. -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] Reference points for incrementals (was: NetBackup 5.1 MP3 Policies)
* bob944 <[EMAIL PROTECTED]> [2006-08-11 00:09]: > > That is correct. You MUST have the fulls and incs in the SAME > > policy for the incs to work. Otherwise, they have nothing to > > compare against and have no way to actually do an incremental > > backup. > > Correct enough, in the context of explaining NetBackup to the neophyte > as you and the previous poster were doing. But the advanced user should > remember that this is the behavior of the cumulative incremental, not > the differential incremental. By definition > o full backs up everything > o cumbacks up everything modified since the last full > o diff backs up everything modified since the last backup That is an important distiction to keep in mind. Digging further, the 5.1 SAG defines a diff as everything since the last full or last diff only, it does not include cumulatives (pg 147). Subtle, but just as important. So many nuances... -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] Aptare vs. NB 6.0 NOM reporting
* Paul Keating <[EMAIL PROTECTED]> [2006-08-11 15:15]: > > > -Original Message- > > From: Ed Wilts [mailto:[EMAIL PROTECTED] > > > Are you running NBU 6.0 perhaps? Because it's seriously busted with > > regards to job status in the Activity Monitor... > > Oh hell no. > I've check the processes from the CLI, and it's definitely not still > running. > ;o) > > I'm working on clearing it now with support. For us, that was an issue with Aptare not properly handling a cross reference of image ids. If you look at the job that's hanging around, drill down and see if there are any related jobids. What you will see is one of the jobs has valid data, but the other one has essentiall no data. Ask them about removing obsolete/old jobs. -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] Aptare vs. NB 6.0 NOM reporting
* Hindle, Greg <[EMAIL PROTECTED]> [2006-08-10 14:18]: > We are looking at reporting software for Netbackup. We are looking at > Storage Console from Aptare. But I understand 6.0 with Nom also has > reporting built in. We are currently running 5.0 mp6 but are planning to > upgrade to 6.0. Is anyone using NOM and the reporting feature? Is it > better than Aptare? Views? Opinions? >From what I saw at Vision '05, NOM is still pretty early in development. I have not looked at it recently, so I can't really speak to strengths/weaknesses. You would like to think that the company that makes the backup software would also have a good front-end for reporting, but that has not been the case so far ;-) I did think it has promise, though. What it did have that Aptare likely will not is more interaction with what happens in the backup environments. It was originally intended to act as a portal interface that at a basic level could replace a lot of the operational functions of the current GUIs. That would be really cool if that pans out. I think in general, most reporting packages will do a better job than NOM because they are supposed to do a better job. Their main focus is reporting, so their reporting is really good. That's not just Aptare's product, either. I think Bocada falls into that same category. There are probably others, but those are the two that seem to be front-runners for NBU reporting. -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] how long does it take to convert catalog ASCII -> binary?
* Vassileff, Glen <[EMAIL PROTECTED]> [2006-08-10 11:20]: > > Hello All > > I am looking for real life experience with NetBackup catalog conversion > ASCII -> binary (15GB) on SUN V490. > How long will it approximately take? Any bad experiences / issues with > such a conversion? Generally, a single image isn't all that long, but there really is no reason to do it. As your older images die, they will be replaced by binary ones moving forward, so it will take only as long as your longest retention for NBU to completely convert to binary on its own. I would guess no more than a couple hours for 15GB. -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] NetBackup 5.1 MP3 Policies
* [EMAIL PROTECTED] <[EMAIL PROTECTED]> [2006-08-10 09:46]: > What you have been told is 100% normal practice. I have zillion of > environments with policies mixing full/inc and different retention levels. > I've seen no environment with separate inc policies. > > Moreover I'm under the impression that separate inc policy always backups > all files (NetBackup detects there is no valid full backup). That is correct. You MUST have the fulls and incs in the SAME policy for the incs to work. Otherwise, they have nothing to compare against and have no way to actually do an incremental backup. -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] exclude list for clients in a specific policy
* DLew97 <[EMAIL PROTECTED]> [2006-08-04 10:18]: > Hi, > > How do I use the Exclude_List to exclude certain files/directories for a > bunch of windows clients in a certain policy? You will need to do this for each client. Under the Client Properties there is a listing of current includes/excludes that you can get as detailed as you need to. You can do it by policy, schedule or globally. What version of NBU are you using? -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] Jobs won't go active
* Ed Wilts <[EMAIL PROTECTED]> [2006-08-02 14:08]: > On Wed, Aug 02, 2006 at 04:34:31PM +0100, Whelan, Patrick wrote: > > I have found that if you run > > bprdreq -rereadconfig; bpschedreq -read_stunits; bpschedreq > > -read_stu_config > > And I've run into a case where doing this killed NetBackup to the point > that no jobs ran. Safer to restart NBU IMO. NBU 6 does allow you to > create STUs on the fly though. Cleaner, not necessarily safer ;-) Seriously, restarting NBU if possible is generally a better idea, especially if all your media servers are dedicated. The problems we have seen usually come from trying to restart NBU and having it lock up during a restart, forcing a reboot of the server. The other problem is if this is happening at a time that you are unable to restart NBU for some reason (like lots of important backups currently running). Most of these are questions because of what political climate your backup system lives in :-) -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] Anyone using VTL
* [EMAIL PROTECTED] <[EMAIL PROTECTED]> [2006-08-01 14:29]: > I guess this begs the question to the audience and the people that have > put in VTL's ... what problems does a VTL solve that D2D can not? > > I think the list so far is: > > 1. Administration is easier > 2. Possibly performance > 3. Compression (1/2 the disk of D2D) > 4. I can use it to stress test a new NBU release > > Anything else? The biggest one that no-one has mentioned yet is the ability to share resources between media servers. With VTL, you can use SSO to allow multiple media servers access to the same resources just like real tape drives. DSSUs are still confined to the one media server that has access to the disk. -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] Browsing from behind firewall?
* Paul Keating <[EMAIL PROTECTED]> [2006-07-31 16:00]: > You need to open bprd (13720) from client to master, for client > initiated actions to work. This is for ALL client-initiated actions, btw. User-Initiated backups, archives, and most database extension-driven backups. -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] Is there a similar email forum for Legato Networker....I know this is sacra religious
* Siano, James C <[EMAIL PROTECTED]> [2006-07-31 12:21]: > Hello. > > I know this is sacra religious but is there a similar email forum for > Legato Networker software. There is no religion involved here. We are all trying to just get through the day and have something to restore from when things go wrong. :-) I'm sure there is something out there, but I have not looked very hard myself. I found this with a quick google: http://listserv.temple.edu/archives/networker.html > I'm now supporting both in my current world and would like to tap a > similar resource. I am sorry. Heterogenous environments suck. -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] NBU 6.0
* Hadrian Baron <[EMAIL PROTECTED]> [2006-07-28 14:54]: > We are going from NB5.0MP5 to 6.0 MPX next month. > > Due to many issues - including corrupt images, massive policy changes, > and most importantly, the perceived difficulty involved with migrating > from Windows 2003 to Solaris 10 environments, I've decided to leave the > existing server online and just zone off the HBAs from our Storagetek. > If I need to restore, zone off the new server and boot the old box. > > Is this too simplistic an approach? We hardly ever do restores from > tape. 99% of restores take place from snapshots on our NetApps gear. The only real issue with that is making sure the client will accept restores from the old server. If you really have only an occasional restore, there is nothing wrong with doing it this way. A clean install is ALWAYS a better way to go if you can deal with leaving the history behind. -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] new client is getting connection refused on 5.1
* Steven L. Sesar <[EMAIL PROTECTED]> [2006-07-21 17:05]: > If you want to connect to the client from the master server GUI, you'll > need port 13722 opened up. If you're trying to backup up through a > firewall, I suggest that you use vnetd, which requires that you open up > port 13724. for basic firewall connectivity, you need: Master -> Client 13782 (bpcd) Client -> Master 13720 (bprd) Client -> Master 13724 (vnetd) And in order to use vnetd, you need to configure the client properties to use it: Host Properties -> Master Server -> Client Attributes -> select vnetd bprd is not specifically necessary for master-initiated backups, but if you want to do User-initiated backups or restores, you will need it. -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] want to give a normal user to monitor netbackup
* Shyingyih Huang \( ?q \) <[EMAIL PROTECTED]> [2006-07-21 10:05]: > Hi: > You can edit the /usr/openv/java/auth.conf > Example : > opmgr ADMIN=AM > > Ps: opmgr is end-user Yes, but if they have access to the Activity Monitor, they have FULL access to it, including the ability to kill active jobs. There is no way to set up "Read Only" access to it. -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] Unix Master, Windows Media server possible?
* Ed Wilts <[EMAIL PROTECTED]> [2006-07-17 06:40]: > On Mon, Jul 17, 2006 at 12:53:51PM +0200, Drunen van, Marcel wrote: > > The whole concept of NetBackup ensures multiplatform operations, which > > is one of the main reasons why I suggest it to my major customers. In > > general my advise is to have the master server on a Windows system, as > > the GUI is much easier to use (and often faster). > > I disagree. Put the master server on a Unix system, and then use the > Windows GUI on a Windows media server or other server. You'll get the > best of both worlds. Or use the CLI ;-) > > The majority of problems I encounter with multi-OS systems is the > > slightly different way UNIX and Windows work with name resolution when > > multiple LAN segments are involved. Make sure your Windows system only > > has one network name, don't try to bind a different name to each NIC, as > > you can easily do on UNIX systems. > > Again, I disagree. Whenever possible, you should use a dedicated backup > interface and you can refer to it by its own name - i.e. > server.prod.foo.com for the front side and server.back.foo.com for the > backup network. Just make sure that NetBackup is configured properly > and you won't have any problems. We've been doing this for years. > > The only place we've been hurt by this is if you have a master/media > server and try to use 2 names for the system at the same time (one for > the master functionality and one for the media functionality). This > worked fine in the 3.x, 4.x, and 5.x days, but broke with 6.0. As long as your server support team knows what they are doing, dual-homing the systems will work. In our environment, the backup team has little control over the client systems and must rely on the server admins to properly handle name resolution and routing. Screw that up, and it gets ugly to troubleshoot in a hurry. A lot of the ideas suggested here do assume a certain amount of technical expertise. Don't ever forget about the "Net" part of NetBackup. :-) -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
Re: [Veritas-bu] Unix Master, Windows Media server possible?
* Ed Wilts <[EMAIL PROTECTED]> [2006-07-15 14:49]: > On Sat, Jul 15, 2006 at 02:15:30PM -0500, David Rock wrote: > > * Siano, James C <[EMAIL PROTECTED]> [2006-07-15 00:17]: > > > Can one have a Unix master and a Windows 200x media server? Wasn't sure > > > if I could have that combo. > > > > Yes, but it is highy recommended to NOT mix UNIX and Windows media > > servers using the same SSO drives. > > We've been doing this since the beginning and it's never caused us any > problems. On what do you base this recommendation? > > > Actually, it is necessary to be able to mix media server types because > > of requirements at times. For example, mounting a Windows filesystem on > > a UNIX media server for off host backing up of SAN snapshots doesn't > > work well. > > Perhaps not, but FlashBackup of a Windows mount point on a Unix media > server does work well. > > > As long as you aren't trying to share tape drives between the different > > media servers, you should be fine. > > I don't see what the issue here. It's supported and it works (for us). We have always been recommended by Veritas support to not do this because of SCSI-release issues between Windows and UNIX systems. The general recommendation to us has always been to avoid mixing Windows and UNIX in SSO :-( -- David Rock [EMAIL PROTECTED] ___ Veritas-bu maillist - Veritas-bu@mailman.eng.auburn.edu http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
