toasters

toasters@lists.teaparty.net

13531 discussions

NDMP and Security
by Sean Hinchey 25 Aug '99

25 Aug '99

We have a UNIX machine running Netbackup. It is initiating backups to locally attached tapes drives on our filer. Can anyone explain how the filer knows it's supposed to allow the UNIX host to initiate backups? I would assume the machine/userid pair would need to be in the hosts.equiv before this would work. However, backups are working without this entry. How does security work with NDMP? -Sean

1 0

NDMP and Security
by Sean Hinchey 25 Aug '99

25 Aug '99

1 0

Re: SIO errors on boot up
by Ono, Gary 25 Aug '99

25 Aug '99

Hello Joe Spainhour, I owuld like to process an RMA for a motherboard replacement on the F330 that was experiencing SIO errors. If could can please provide me with the serial number to the filer and verify the shipping address I will process the RMA. I look forward to your response. Gary REQUEST RMA's ONLINE http://now.netapp.com/cgi-bin/initiaterma ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Gary Ono 495 East Java Drive Network Appliance Sunnyvale, CA 94089 Customer Support Rep. Phone: 408.822.4761 www.netapp.com Fax: 408.822.4408 gono(a)netapp.com ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Tech Support Hotline: 1.888.4.NETAPP

1 0

what's a good NFS write speed in MB/sec?
by Matt Harrington 25 Aug '99

25 Aug '99

I'm evaluating a NAS filer similar to the ones sold by NetApp and Auspex. This particular model is made by Nitech in Irvine, CA. These NFS vendors speak about benchmarks in ops/sec. It's more intuitive for me to think of benchmarks in terms of MB/sec throughputs for a large file write or read. My quick-and-dirty tests show that I can write at 1.8 MByte/sec over a 100Mbit network using NFS v3. I can FTP at about 7 MByte/sec. My question is: do these numbers seem reasonable? If not, what aspects of NFS should I tweak? ---matt p.s. methods: to make a 100MB file: dd if=/dev/zero of=/tmp/bigfile bs=1024 count=102400 time cp /tmp/bigfile /home/bigfile (where /home is the NFS server).

3 3

Netapp CIFS as part of DFS tree?
by Brian Tao 25 Aug '99

25 Aug '99

I got this question from one of our NT guys... can a CIFS-enabled Netapp participate in an NT DFS tree? From what I understand, NT's DFS looks and acts like UNIX-style directory mounts (i.e., instead of assigning a drive letter to a shared drive, the drive appears under an existing directory). -- Brian Tao (BT300, taob(a)risc.org) "Though this be madness, yet there is method in't"

4 4

RE: Copying small directory on netapp
by Muhlestein, Mark 24 Aug '99

24 Aug '99

You can use 'scopy.exe' from the NT side. If you run it as Administrator it should preserve all UNIX and NT metadata, including symlinks, DOS bits, UNIX permissions, etc. Mark Muhlestein -- mmm(a)netapp.com -----Original Message----- From: Stephen C. Losen [mailto:scl@sasha.acc.virginia.edu] Sent: Tuesday, August 24, 1999 8:23 AM To: toasters(a)mathworks.com Subject: Copying small directory on netapp We are using our netapps for home directories in a mixed NFS and CIFS environment. We have a prototype new user directory that we set up with some unix "dot" files, a public_html directory, etc. Most of our CIFS users never touch unix, so from CIFS we make the unix dot files hidden. We do this in the hope that they won't delete them and then later on give unix a try and have problems when their account doesn't behave as documented. We use the netapp dump program to dump the prototype directory to a disk file. We create new accounts from unix. We make a new user directory and populate it with netapp restore. Then we chown all the files to the new user. This works great because it preserves the "hidden" attribute on all the "dot" files. Unfortunately, the netapp dump utility writes all the inode information for the entire filesystem to the dump file. So our little prototype directory, which tars into a file of about 20K, dumps into a 2.5M file. This really slows down bulk user adds, because it takes several seconds for each restore to run. Not a surprise when you have to rsh over 2.5M. We added 4700 accounts for our new incoming students and it took about a day -- almost all of the time taken up by restore. And yes, we dump once to file, and do multiple restores from it. I suppose I could build a very small and empty netapp filesystem and put the prototype in there and dump it. That would certainly reduce the size of the dump, but that's an impractial waste of resources. So here are my questions: 1) Is there some way other than dump/restore to copy a directory on a netapp that preserves all NFS and CIFS attributes? 2) We are running 5.2.2P1 and I don't see any dump flags that avoid including all the inode info in the dump. Am I missing something? 3) Is there a way from unix to set the CIFS hidden attribute? If so, I would just use tar and set the attribute. Steve Losen scl(a)virginia.edu phone: 804-924-0640 University of Virginia ITC Unix Support

1 0

RE: Netapp CIFS as part of DFS tree?
by Muhlestein, Mark 24 Aug '99

24 Aug '99

Any CIFS client can be a destination in an NT DFS setup because of the way it works. If a path is to be rerouted, the client gets a special error, then asks the server for the new path. The client then issues a separate connection to the destination share, which can be on any CIFS server anywhere, including the filer. Mark Muhlestein -----Original Message----- From: Brian Tao [mailto:taob@risc.org] Sent: Tuesday, August 24, 1999 8:08 AM To: toasters(a)mathworks.com Subject: Netapp CIFS as part of DFS tree? I got this question from one of our NT guys... can a CIFS-enabled Netapp participate in an NT DFS tree? From what I understand, NT's DFS looks and acts like UNIX-style directory mounts (i.e., instead of assigning a drive letter to a shared drive, the drive appears under an existing directory). -- Brian Tao (BT300, taob(a)risc.org) "Though this be madness, yet there is method in't"

2 1

Re: SIO error on boot up.
by Guy Lauterbach 24 Aug '99

24 Aug '99

At 03:37 PM 8/24/99 -0400, Joe Spainhour wrote: >We are getting SIO errors on an F330 while it is doing its POST. >It then halts while trying to load the OS. >Any ideas on this? This is an error in a pretty fundamental part. Basically, the message means that the PCI/ISA bridge chip (SIO - Intel 82378ZB) has a problem. Since this is just about the very first test in a filer at power-on, it could be indicative of a number things, such as clocks, interrupts, or PCI operation. There is no workaround. I think you'll have to RMA your motherboard. rgl R. Guy Lauterbach Manager, Firmware & Diagnostics Network Appliance, Inc.

1 0

No autosupport mail on reboot
by Graham C. Knight 24 Aug '99

24 Aug '99

Has anyone encountered this type of problem? After upgrading to 5.2.3 none of our systems send autosupport mail on reboot. I can issue an "options autosupport.doit test" command and autosupport mail is sent, but nothing on reboot. There is nothing in the messages file that indicates an error, also there is no message that autosupport has been sent. Here's what the autosupport lines in rc look like: autosupport.mailhost mailhost options autosupport.to autosupport@netapp.com,nac-admin@ast.lmco.com options autosupport.from nac-admin-reply(a)ast.lmco.com I don't see a whole lot of ways to debug this problem. Any tips would be greatly appreciated. Thx, Graham

2 1

SIO error on boot up.
by Joe Spainhour 24 Aug '99

24 Aug '99

We are getting SIO errors on an F330 while it is doing its POST. It then halts while trying to load the OS. Any ideas on this? Joseph

1 0

RE: Netapp CIFS as part of DFS tree?
by Dawn Hamel 24 Aug '99

24 Aug '99

We here at SAS do this all the time. You can do DFS mapping via the command line. Something like: dfscmd /map \\sashq\u\userid \\netapp_name\share_name\userid And the user will just need to go down to \\sashq\u\userid. They do not need to know where this area resids. MS DFS hides this from the user (Just like AFS/DFS does in the Unix world) Regards, Dawn -----Original Message----- From: Brian Tao [mailto:taob@risc.org] Sent: Tuesday, August 24, 1999 11:08 AM To: toasters(a)mathworks.com Subject: Netapp CIFS as part of DFS tree? I got this question from one of our NT guys... can a CIFS-enabled Netapp participate in an NT DFS tree? From what I understand, NT's DFS looks and acts like UNIX-style directory mounts (i.e., instead of assigning a drive letter to a shared drive, the drive appears under an existing directory). -- Brian Tao (BT300, taob(a)risc.org) "Though this be madness, yet there is method in't"

1 0

Re: No autosupport mail on reboot
by sirbruce＠ix.netcom.com 24 Aug '99

24 Aug '99

On 08/24/99 08:22:24 you wrote: >I don't see a whole lot of ways to debug this problem. Any >tips would be greatly appreciated. If I recall correctly, autosupport mail is not sent when booting off of floppy. So if you booted off of floppy to do the upgrade, you won't get an autosupport. One way to test if this is the problem is to download the upgrade to the disks, pop out the floppy, and then reboot off the disk image. Bruce

1 0

RE: Gigabit Ethernet
by Jonas Jonsson 22 Aug '99

22 Aug '99

I would recommend that you turn off auto-neg and do manual set-up on the interface. We have used Extreme and Netapp Gb eth with hard coded settings. /// Best regards, Jonas -----Original Message----- From: Jay Querusio [mailto:jq@opensystems.com] Sent: 20 August 1999 21:16 To: toasters(a)mathworks.com Subject: Gigabit Ethernet Has anyone ever experienced any problems with a GB Ethernet card on a Netapp connected to an Extreme Black Diamond switch? The card seems to come up fine but after a period of time the Netapp becomes unreachable. The amber and green lights on the back of the card seem to stay a dim steady light when this happens. All the interface statistics seem to be fine with no collisions. The mtu on the switch is set to 1500. The only difference we can find is that the filer labels the card as 1000fx while the switch port can only do sx or lx. Any one know what could be going on or have any ideas? Thanks in advance, Jay

2 1

Re: ndmpd status
by sirbruce＠ix.netcom.com 20 Aug '99

20 Aug '99

On 08/20/99 12:59:19 you wrote: >My problem has been solved though, because someone >pointed out to me that if the filer is in rc_toggle_basic >mode you can rsh to ndmpd. Many of the commands that are only rsh-able via the rc_toggle_basic are such because they are not necessarily *safe* to be rsh-ed commands. So I generally would not think of this as solving your problem, unless you're willing to live with possible crashes and other undefined behavior as a result under the right conditions. The true solution is to upgrade to a release that supports that command through rsh. Bruce

1 0

Gigabit Ethernet
by Jay Querusio 20 Aug '99

20 Aug '99

Has anyone ever experienced any problems with a GB Ethernet card on a Netapp connected to an Extreme Black Diamond switch? The card seems to come up fine but after a period of time the Netapp becomes unreachable. The amber and green lights on the back of the card seem to stay a dim steady light when this happens. All the interface statistics seem to be fine with no collisions. The mtu on the switch is set to 1500. The only difference we can find is that the filer labels the card as 1000fx while the switch port can only do sx or lx. Any one know what could be going on or have any ideas? Thanks in advance, Jay

1 0

ndmpd status
by Graham Knight 20 Aug '99

20 Aug '99

I was quite suprised that you cannot do a "rsh toaster ndmpd status". Does anyone know of any tricks to get the status of ndmpd connections from a Unix shell? I want to make sure connections are available before i launch ndmpcopy processes. Along the same lines - is it possible to increase the amount of concurrent ndmp sessions beyond the default of 5? Thanks, Graham

2 2

2Gig filesize limit?
by Rob Windsor 20 Aug '99

20 Aug '99

Hello, We're running into a bug where the Synopsys tools blow up when the simulation datafile size reaches 2Gig. It appears that Solaris 2.6 is fine with 'largefile' as long as the application is appropriately set up for such. Before I send the developer off to pound on Synopsys about the problem, does Data OnTapp 5.2.1 have a filesize limitation? I seem to remember reading "not since v4.n" (and a given value of `n'), but now I can't find that reference. -- Rob ---------------------------------------- Rob Windsor E-Mail - mailto:windsor@adc.com Senior Unix Systems Administrator Voice - phone:972-680-6919 Computer Services Fax - phone:972-680-0370 Business Broadband Group __o ADC Telecommunications _`\<,_ Richardson, TX 75082 (_)/ (_)

3 3

RE: ndmpd status
by Melvin, Grant 20 Aug '99

20 Aug '99

Hi Graham, This is resolved in Data ONTAP 5.3 and higher. ndmpd is rsh'able. In Data ONTAP 5.3 we ported the NDMP server to the JVM on the filer. One of the reasons to allow more than 5 NDMP concurrent sessions. We have now set an arbitary limit of 1000. Cheers, Grant -----Original Message----- From: Graham Knight [mailto:grahamk@ast.lmco.com] Sent: Friday, August 20, 1999 11:10 AM To: toasters(a)mathworks.com Subject: ndmpd status I was quite suprised that you cannot do a "rsh toaster ndmpd status". Does anyone know of any tricks to get the status of ndmpd connections from a Unix shell? I want to make sure connections are available before i launch ndmpcopy processes. Along the same lines - is it possible to increase the amount of concurrent ndmp sessions beyond the default of 5? Thanks, Graham

1 0

tech support waits
by Todd C. Merrill 19 Aug '99

19 Aug '99

I am getting frustrated with the response time from tech support. The engineers who take the calls are not responding to my requests for status updates, for days even. Not even a peep, such as, "We're still working on it." I have not tried the "comments" section from the call log on NOW, but I don't have faith in that if supposed *direct* communication to the engineer isn't working. I have started calling the 888-4NETAPP number, and waiting on hold for 20-30+ minutes. There is no way to quickly just say, "Have the engineer call me, *please*"; you have to wait...and wait....and wait.... I waited over 20 minutes the other day. Frustrated, I hung up, redialed, and entered the option for "the filer is not serving data," hoping I could get someone quickly. Hah. 18 minutes later someone answered the call. I asked the engineer if today was a busy day; the reply was, "Not really." I don't know about other people, but if this had been an actual emergency and our filer was down, waiting 20 minutes is unacceptable. There were threads a while back about response and wait times. Are others still experiencing these interminable delays? In defense of NetApp, though, I did make a call at 6 AM (EDT) last Friday, and got someone immediately from their third-shift in Europe. Yea! That is much improved! (Now, if I could only get a hold of that engineer *again*...) Until next time... The Mathworks, Inc. 508-647-7000 x7792 3 Apple Hill Drive, Natick, MA 01760-2098 508-647-7001 FAX tmerrill(a)mathworks.com http://www.mathworks.com ---

4 3

RE: HELP TO DECIDED THE ONTAP VERSION TO USE WITH BUDTOOL
by Melvin, Grant 13 Aug '99

13 Aug '99

Hi Alex, I received good news last night that BudTool 4.6.1 is now certified with Data ONTAP 5.2.3: NetApp Release 5.2.3 has been tested, and certified with BudTool 4.6.1. Local, Remote, and 3 party configurations were tested. Legato havn't updated their external web page yet at: http://www.iguard.com/products/certmatrix.html since the product marketing manager is out today. It should be updated next week sometime. Legato will be certifying 5.3.2D1 next. Cheers, Grant P.S. We have moved our internal version of the NetApp/VERITAS (NetBackup) software certification matrix to: http://now.netapp.com/knowledge/contents/TIP/TIP_605.shtml These URLs should provide you with the information that you require. -----Original Message----- From: Alexandre Gomes [mailto:r7043c@email.sps.mot.com] Sent: Friday, August 13, 1999 10:58 AM To: toasters(a)mathworks.com Subject: HELP TO DECIDED THE ONTAP VERSION TO USE WITH BUDTOOL Hello, Could some one help me decided what version of Network appliance ONTAP is compatible with budtool version 4.5 and 4.6. It it the version more recently of budtool ??. If not, what is the version more recent od budtool and what version of ONTAP is compatible with it ? The version that I'm using in my Network Appliance F230 is the ONTAP 4_3_4 . We intend to install the version 5_2_1. But I would like to Know the information above ... Someone Know how many time it takes to be installed a shelf disk and do the actualization of the version to 5_2_1 ?? Thanks very much, Regards, Alex

1 0

HELP TO DECIDED THE ONTAP VERSION TO USE WITH BUDTOOL
by Alexandre Gomes 13 Aug '99

13 Aug '99

Hello, Could some one help me decided what version of Network appliance ONTAP is compatible with budtool version 4.5 and 4.6. It it the version more recently of budtool ??. If not, what is the version more recent od budtool and what version of ONTAP is compatible with it ? The version that I'm using in my Network Appliance F230 is the ONTAP 4_3_4 . We intend to install the version 5_2_1. But I would like to Know the information above ... Someone Know how many time it takes to be installed a shelf disk and do the actualization of the version to 5_2_1 ?? Thanks very much, Regards, Alex

1 0

Re: The Promise Of SAN ???.....EMC answer....
by sirbruce＠ix.netcom.com 13 Aug '99

13 Aug '99

>No company left the FibreAlliance. This was a mis-interpretation of >only one reporter. Note that no other press reported the same story. Whether or not the story is true, the "only one press reported it" line is hardly a defense. Remember Woodward and Bernstein? >Finally, I have seen lots of presentations by analysts about NAS and >SAN and what is stated below is really focused on the point they make >of where do NAS and SAN meet. If you need to have file sharing, then >use NAS but even the NAS server will need the benefits of SAN >(connectivity, distance, scaleability, etc). So the analysts position >NAS as using SAN on their backend. This was clearly the direction 2 years ago, and if it wasn't for SAN vendor fud more customers would realize this. >There are still, and will continue >to be, a large installed set of application that do not use NAS and >they will use SAN directly. There will continue to be a far larger installed set of applications that use only NAS and not SAN. In fact, very few applications use *only* SAN directly, because that would largely be useless; the data still has to get to the end-user to be useful, even if it's through a secondary application, and that user is very likely to be connected over a network. Bruce

1 0

Re: NFS lock tools
by sirbruce＠ix.netcom.com 13 Aug '99

13 Aug '99

On 08/13/99 00:55:17 you wrote: >Vic Abell, the author of the o-so-useful lsof program (which knows >about locks in some cases), also has a program called lslk for listing >existing locks. Both lsof and lslk can be found at ><ftp://vic.cc.purdue.edu:/pub/tools/unix/>. I've communicated with >Mr. Abell, and he does not know of any specific tool to do what we've >been discussing. I'd like to second the recommendation of lsof, which is essentially the second incarnation of the ofiles command. I did some work on it for Vic as an undergrad. As the the locking issue, I had always found simply restarting lockd and statd on the client were sufficient, without the procedure of locking another file in-between. But perhaps it depends on the client and the filer OS version. Bruce

1 0

The Promise Of SAN ???.....EMC answer....
by Savini, Patrice 13 Aug '99

13 Aug '99

Here is EMC answer...Regards, Patrice ------- Forwarded Message Follows ------- From: "blumenau, steven" <blumenau_steven(a)emc.com> To: "'tzelnic(a)emc.com'" <tzelnic(a)emc.com> Copies to: "ofer, erez" <erez(a)emc.com> Subject: RE: (Fwd) FW: The Promise Of SAN ???..... Date sent: Thu, 12 Aug 1999 12:24:17 -0400 Percy, No company left the FibreAlliance. This was a mis-interpretation of only one reporter. Note that no other press reported the same story. The FibreAlliance is focused on managing SANs. The OSFI is focused on working on interoperability of the protocol for switches. Specifically, how do two different vendors switches talk to each other over eports. All the work that the FibreAlliance does is submitted to standards bodies (ANSI, IETF, DMTF, etc). So there is no reason that they can be considered proprietary. For example, the MIB is currently a proposed draft in IETF and changes to the protocols have been submitted to ANSI. The quote below from Roger (Reich) was mis-quoted out of context. Roger even sent in a letter to unigram about this and it was published. As to all the groups working on standards, it turns out that we are all working together very well. The FibreAlliance is focused on SAN management at the lower levels of protocols, APIs and MIBs. It turns out the SNIA has adopted the MIB that we submitted to IETF and have proposed changes which have been incorporated into it. Finally, I have seen lots of presentations by analysts about NAS and SAN and what is stated below is really focused on the point they make of where do NAS and SAN meet. If you need to have file sharing, then use NAS but even the NAS server will need the benefits of SAN (connectivity, distance, scaleability, etc). So the analysts position NAS as using SAN on their backend. There are still, and will continue to be, a large installed set of application that do not use NAS and they will use SAN directly. Steve > -----Original Message----- > From: Percy Tzelnic [mailto:tzelnic@emc.com] > Sent: Thursday, August 12, 1999 10:06 AM > To: blumenau_steve(a)emc.com > Cc: erez(a)emc.com > Subject: (Fwd) FW: The Promise Of SAN ???..... > > > What's the answer? > Thanks, > Percy > > ------- Forwarded Message Follows ------- > From: "Savini, Patrice" <p-savini(a)ti.com> > To: "'Tzelnic'" <tzelnic(a)emc.com> > Subject: FW: The Promise Of SAN ???..... > Date sent: Thu, 12 Aug 1999 14:39:25 +0200 > > Any comment ? Patrice > > ************************************************************** > ********* > > EMC's Fibre Alliance Woes and the state of SAN Interoperability > On 7/23 Jo Maitland announced in UNIGRAM, that more than half of the > Fibre > Channel switch manufacturers (McData, Brocade, Ancor, Gadzoox > and Vixel) > > defected from EMC's Fibre Alliance, "fed-up with its slow pace on > standardization". These 5 vendors have created their own, independent > group, > (the Open Standards Fabric Initiative - OSFI) charged with settings > standards of interoperability between their own products and between > their > products and third party products. > > Key points: > 1. Standards exist for interoperability between switches, and third > party > products, but each vendor was using their own interpretation of those > standards. This led to a lack of interoperability between different > manufacturers switch products and the need for customers to purchase > proprietary SAN solutions - something they are not interested > in doing. > This > also led to lack of plug-and-play interoperability between > switches and > third party products (such as the HBAs needed in servers to connect to > the > SAN) - in turn forcing vendors of SANs to go through exhaustive > interoperability testing every time a new product revision > was released > or a > new product line was added to their suite of SAN products. > > 2. EMC set up the Fibre Alliance in February with great fanfare, > splitting > off from the SNIA to develop interoperability between components. > Notably > absent from the group were any other storage vendors such as IBM or > Hitachi. > As a result it was seen by many in the industry as a way for EMC to > force > it's own standards on the rest of the Fibre and SAN > community. An effort > > that has obviously failed - at least for now. > > 3. Despite the fact that the OSFI has stated that it is a subgroup > within > the FibreAlliance, EMC has stated that they will monitor the group to > see if > their work conflicts with what the Fibre Alliance is doing - in other > words > EMC will do what they're doing, without input or feedback > from the OSFI, > > leading yet again to dual (and dueling!) standards. An indication of > this is > contained with the report: "One analyst on the call said that > working on > > device level interoperability is fine but without an overall framework > for > managing those devices, the storage area network is still way off. A > spokesperson for McData, a member of OSFI said, "We have to > crawl before > we > can walk." " > > 4. A third group working on standards is the SNIA (Storage Network > Industry > Association) with over 100 members. A quote from the SNIA and from a > member > of the OSFI sums up the situation with multiple standards committees > working > on the same issues but in opposite directions. SNIA chairman, Roger > Reich, > director of Compaq's network storage division said of the new > group: "I > don't know where they are going or what they are doing, but > if they need > > backing in the form of money or talent we would be happy to speak to > them." > He added that the Fibre Alliance was a closed, profit-making > group with > a > different agenda to the SNIA. In response to Reich, Greg Reyes, CEO of > Brocade and a member of the OSFI said: "We are not interested in his > offer." > > Summary > 1. Despite their protestations to the contrary, EMC is > proprietary, will > > continue to be proprietary and does not wish to work on a global scale > with > other vendors to open up interoperability of servers, storage devices > and > interconnect products such as switches, hubs and HBAs. Customer > purchasing > EMC storage devices or SAN implementations will be locking themselves > into > the high price leader and into a proprietary solution. > 2. This lends further credence to ****** stated vision (backed up by > analysts) of NAS as the "Intelligent Gateway" to the SAN, and that NAS > should be implemented today with an eye toward tomorrows SAN > implementation > (i.e. SAN should not be implemented today). Install SAN today at your > own > risk, if for no other reason than a lack of standards and a lack of a > current methodology of moving toward a single set of globally accepted > standards. > > >

2 1

RE: NetGroup issue..last time
by Mohler, Jeff 13 Aug '99

13 Aug '99

Yep, many of you bent me in the right direction, actually a sudden vision of clarity about11 last night got me to the same place, but this AM's Email confirms. I _really_ must admit, that Toasters isnt line any other technical mailing list where "Duh, its in the manual" is the most common answer. Thanks! -----Original Message----- From: Stephen C Woods [mailto:scw@seas.ucla.edu] Sent: Thursday, August 12, 1999 6:03 PM To: steveb(a)sh-gpl.ti.com; jeff.mohler(a)wilcom.com Subject: Re: NetGroup issue..last time Steve, If by 'netaccess' you really ment root access you are correct. Granting root access requires a specific list of hosts. <scw> > From steveb > According to my docs. You can't give netgroups netaccess. Just specific hosts. > Although I am only running version 3.1. this may have changed in later versions. > According to Mohler, Jeff : > :-) > :-) Ok..sendmail ate my brain today, so a simple netgroup insertion..NO problem. > :-) > :-) My /etc/netgroup file on the filer: > :-) > :-) dgrmouse# more netgroup > :-) News-Feeders (news-feeder2-qfe1.wiltec.twc.com,,) > :-) (news-feeder2-qfe0.wiltec.t > :-) wc.com,,) (news-feeder-qfe0.wcg.net,,) (news-feeder-qfe1.wcg.net,,) > :-) News-Readers (news-reader-qfe0.wcg.net,,) (news-reader-qfe1.wcg.net,,) > :-) (news- > :-) reader3-qfe1.wcg.net,,) (news-reader3-qfe0.wcg.net,,) > :-) > :-) All hosts resolve, and are also entered in /etc/hosts on the filer. > :-) > :-) My attempted export: > :-) > :-) /vol/NetMaint -access=News-Feeders,root=News-Feeders > :-) > :-) Output: > :-) > :-) ratbert> exportfs -a > :-) export: Can't find address for host News-Feeders > :-) > :-) > :-) ? > :-) > :-) > :-) Everything Ive looked at, "feels" ok, looks OK, but the filer dont think > :-) it's ok. Netgroup is told to look in 'files' initially in Name > :-) Services/DNS as well. > :-) ----- Stephen C. Woods; UCLA SEASnet; 2567 Boelter hall; LA CA 90095; (310)-825-8614 Finger for public key scw(a)cirrus.seas.ucla.edu,Internet mail:scw@SEAS.UCLA.EDU

1 0

Jump to page:

2026

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

2001

2000

1999

1998

1997

toasters