toasters May 2008

toasters@lists.teaparty.net

51 participants
41 discussions

VIF weirdness
by Page, Jeremy 16 May '08

16 May '08

I am having a great deal of difficulty getting my filer's 10g interfaces connected to a pair of 3750E switches I am trying to use for my storage network's backbone. I can inconsistently ping other objects on the network (there really is nothing there yet, we are just getting started, but I do have a host and an RLM card just for troubleshooting purposes), but they cannot ping each other. From a remote host (172.1.2.5) I can ping everything but array1. If I bring down either of the two stacked switches (which brings down one port of each VIF member pairs) everything works. array01> ping 172.1.0.1 172.1.0.1 is alive array01> ping 172.1.0.2 array01> ping 172.1.0.4 ping: wrote 172.1.0.4 64 chars, error=Host is down ping: wrote 172.1.0.4 64 chars, error=Host is down array01> Thu May 15 13:28:33 EDT [gvr-array01: nis_worker_0:info]: Local NIS group update successful. array01> ping 172.1.2.5 172.1.2.5 is alive array02> ping 172.1.0.1 array02> ping 172.1.0.4 172.1.0.4 is alive array02> ping 172.1.0.4 172.1.0.4 is alive array02> ping 172.1.0.250 172.1.0.250 is alive array02> ping 172.1.2.5 no answer from 172.1.2.5 array01> ifconfig -a e0a: flags=948043<UP,BROADCAST,RUNNING,MULTICAST,TCPCKSUM> mtu 1500 ether 02:a0:98:08:22:b7 (auto-1000t-fd-up) flowcontrol full trunked lan0 e0b: flags=108042<BROADCAST,RUNNING,MULTICAST,TCPCKSUM> mtu 1500 ether 00:a0:98:08:22:b6 (auto-unknown-cfg_down) flowcontrol full e0c: flags=948043<UP,BROADCAST,RUNNING,MULTICAST,TCPCKSUM> mtu 1500 ether 02:a0:98:08:22:b7 (auto-1000t-fd-up) flowcontrol full trunked lan0 e0d: flags=108042<BROADCAST,RUNNING,MULTICAST,TCPCKSUM> mtu 1500 ether 00:a0:98:08:22:b4 (auto-unknown-cfg_down) flowcontrol full e2a: flags=948043<UP,BROADCAST,RUNNING,MULTICAST,TCPCKSUM> mtu 1500 ether 02:a0:98:08:22:b6 (auto-10g_sr-fd-up) flowcontrol full trunked tgif e2b: flags=948043<UP,BROADCAST,RUNNING,MULTICAST,TCPCKSUM> mtu 1500 ether 02:a0:98:08:22:b6 (auto-10g_sr-fd-up) flowcontrol full trunked tgif lo: flags=1948049<UP,LOOPBACK,RUNNING,MULTICAST,TCPCKSUM> mtu 8160 inet 127.0.0.1 netmask 0xff000000 broadcast 127.0.0.1 ether 00:00:00:00:00:00 (VIA Provider) lan0: flags=948043<UP,BROADCAST,RUNNING,MULTICAST,TCPCKSUM> mtu 1500 inet 10.28.17.213 netmask 0xffffff00 broadcast 10.28.17.255 partner 10.28.17.214 (not in use) ether 02:a0:98:08:22:b7 (Enabled virtual interface) tgif: flags=4948043<UP,BROADCAST,RUNNING,MULTICAST,TCPCKSUM,NOWINS> mtu 1500 inet 172.1.0.1 netmask 0xffff0000 broadcast 172.1.255.255 partner 172.1.0.2 (not in use) ether 02:a0:98:08:22:b6 (Enabled virtual interface) nfo enabled array02>ifconfig -a e0a: flags=948043<UP,BROADCAST,RUNNING,MULTICAST,TCPCKSUM> mtu 1500 ether 02:a0:98:0c:1a:37 (auto-1000t-fd-up) flowcontrol full trunked lan0 e0b: flags=108042<BROADCAST,RUNNING,MULTICAST,TCPCKSUM> mtu 1500 ether 00:a0:98:0c:1a:36 (auto-unknown-cfg_down) flowcontrol full e0c: flags=948043<UP,BROADCAST,RUNNING,MULTICAST,TCPCKSUM> mtu 1500 ether 02:a0:98:0c:1a:37 (auto-1000t-fd-up) flowcontrol full trunked lan0 e0d: flags=108042<BROADCAST,RUNNING,MULTICAST,TCPCKSUM> mtu 1500 ether 00:a0:98:0c:1a:34 (auto-unknown-cfg_down) flowcontrol full e2a: flags=948043<UP,BROADCAST,RUNNING,MULTICAST,TCPCKSUM> mtu 1500 ether 02:a0:98:0c:1a:36 (auto-10g_sr-fd-up) flowcontrol full trunked tgif e2b: flags=948043<UP,BROADCAST,RUNNING,MULTICAST,TCPCKSUM> mtu 1500 ether 02:a0:98:0c:1a:36 (auto-10g_sr-fd-up) flowcontrol full trunked tgif lo: flags=1948049<UP,LOOPBACK,RUNNING,MULTICAST,TCPCKSUM> mtu 8160 inet 127.0.0.1 netmask 0xff000000 broadcast 127.0.0.1 ether 00:00:00:00:00:00 (VIA Provider) lan0: flags=948043<UP,BROADCAST,RUNNING,MULTICAST,TCPCKSUM> mtu 1500 inet 10.28.17.214 netmask 0xffffff00 broadcast 10.28.17.255 partner 10.28.17.213 (not in use) ether 02:a0:98:0c:1a:37 (Enabled virtual interface) tgif: flags=4948043<UP,BROADCAST,RUNNING,MULTICAST,TCPCKSUM,NOWINS> mtu 1500 inet 172.1.0.2 netmask 0xffff0000 broadcast 172.1.255.255 partner 172.1.0.1 (not in use) ether 02:a0:98:0c:1a:36 (Enabled virtual interface) nfo enabled This message (including any attachments) contains confidential and/or proprietary information intended only for the addressee. Any unauthorized disclosure, copying, distribution or reliance on the contents of this information is strictly prohibited and may constitute a violation of law. If you are not the intended recipient, please notify the sender immediately by responding to this e-mail, and delete the message from your system. If you have any questions about this e-mail please notify the sender immediately.

4 3

Multistore Setup
by Cristian Rojas 15 May '08

15 May '08

Hi list, this is my env: -FAS960 -OnTAP 7.2.3 -2 Aggr -Installed Multistore License -Created vfilers: lgsstpnas02> vfiler status vfiler0 running excvfiler1 running But cant ping the ip of the vfiler assigned to e6b. Here my network interfaces: e0 Up 192.168.215.95 Modify Up Down e5b Down 10.223.61.90 Modify Up Down e6b Up 10.223.61.91 Modify Up Down e7b Down -- Modify Up Down e9b Down -- Modify Up Down vifp32 Up 10.223.59.111 Modify Up Down RemoveTrunk - e5a Up -- Modify - e6a Up -- Modify vifpriv Up 192.168.56.4 Modify Up Down RemoveTrunk - e7a Up -- Modify - e9a Up -- Modify lo Up 127.0.0.1 The two cisco 3750 on top are trunked so I should be able to reach any vlan in my network. Any hint anyone? Thanks! Cris ____________________________________________________________________________________ Be a better friend, newshound, and know-it-all with Yahoo! Mobile. Try it now. http://mobile.yahoo.com/;_ylt=Ahu06i62sR8HDtDypao8Wcj9tAcJ

3 3

RE: VIF weirdness
by Page, Jeremy 15 May '08

15 May '08

They are 10 gig cards, they are only in a VIF with themselves (no other cards). It's a 3070 with 2 copper 100/1000 2 FC gig and 2 FC 10g connections. The 2 1000 are on our production network as lan0 VIF for CIFS/NFS, the fibre gig is not used and the two 10g FC are VIF tgif on the storage only network (no routing etc, private for storage devices and the hosts using them only). They are plugged into stacked 3750e boxes, I tried putting them on the same switch just to make sure it was not a stacking thing with no real change. The oddness is that occasionally a single packet will get through, and if I bring the interfaces up and down it may change who can get to what. ________________________________ From: Webster, Stetson [mailto:Stetson.Webster@netapp.com] Sent: Thursday, May 15, 2008 2:13 PM To: Page, Jeremy; toasters(a)mathworks.com Subject: RE: VIF weirdness Are these 10g TOE cards? Are they in a VIF with other 'non-TOE' cards? Stetson M. Webster Onsite Professional Services Engineer PS - North Amer. - East NetApp 919.250.0052 Mobile Stetson.Webster(a)netapp.com www.netapp.com <http://www.netapp.com/> ________________________________ From: Page, Jeremy [mailto:jeremy.page@gilbarco.com] Sent: Thursday, May 15, 2008 1:32 PM To: toasters(a)mathworks.com Subject: VIF weirdness I am having a great deal of difficulty getting my filer's 10g interfaces connected to a pair of 3750E switches I am trying to use for my storage network's backbone. I can inconsistently ping other objects on the network (there really is nothing there yet, we are just getting started, but I do have a host and an RLM card just for troubleshooting purposes), but they cannot ping each other. >From a remote host (172.1.2.5) I can ping everything but array1. If I bring down either of the two stacked switches (which brings down one port of each VIF member pairs) everything works. array01> ping 172.1.0.1 172.1.0.1 is alive array01> ping 172.1.0.2 array01> ping 172.1.0.4 ping: wrote 172.1.0.4 64 chars, error=Host is down ping: wrote 172.1.0.4 64 chars, error=Host is down array01> Thu May 15 13:28:33 EDT [gvr-array01: nis_worker_0:info]: Local NIS group update successful. array01> ping 172.1.2.5 172.1.2.5 is alive array02> ping 172.1.0.1 array02> ping 172.1.0.4 172.1.0.4 is alive array02> ping 172.1.0.4 172.1.0.4 is alive array02> ping 172.1.0.250 172.1.0.250 is alive array02> ping 172.1.2.5 no answer from 172.1.2.5 array01> ifconfig -a e0a: flags=948043<UP,BROADCAST,RUNNING,MULTICAST,TCPCKSUM> mtu 1500 ether 02:a0:98:08:22:b7 (auto-1000t-fd-up) flowcontrol full trunked lan0 e0b: flags=108042<BROADCAST,RUNNING,MULTICAST,TCPCKSUM> mtu 1500 ether 00:a0:98:08:22:b6 (auto-unknown-cfg_down) flowcontrol full e0c: flags=948043<UP,BROADCAST,RUNNING,MULTICAST,TCPCKSUM> mtu 1500 ether 02:a0:98:08:22:b7 (auto-1000t-fd-up) flowcontrol full trunked lan0 e0d: flags=108042<BROADCAST,RUNNING,MULTICAST,TCPCKSUM> mtu 1500 ether 00:a0:98:08:22:b4 (auto-unknown-cfg_down) flowcontrol full e2a: flags=948043<UP,BROADCAST,RUNNING,MULTICAST,TCPCKSUM> mtu 1500 ether 02:a0:98:08:22:b6 (auto-10g_sr-fd-up) flowcontrol full trunked tgif e2b: flags=948043<UP,BROADCAST,RUNNING,MULTICAST,TCPCKSUM> mtu 1500 ether 02:a0:98:08:22:b6 (auto-10g_sr-fd-up) flowcontrol full trunked tgif lo: flags=1948049<UP,LOOPBACK,RUNNING,MULTICAST,TCPCKSUM> mtu 8160 inet 127.0.0.1 netmask 0xff000000 broadcast 127.0.0.1 ether 00:00:00:00:00:00 (VIA Provider) lan0: flags=948043<UP,BROADCAST,RUNNING,MULTICAST,TCPCKSUM> mtu 1500 inet 10.28.17.213 netmask 0xffffff00 broadcast 10.28.17.255 partner 10.28.17.214 (not in use) ether 02:a0:98:08:22:b7 (Enabled virtual interface) tgif: flags=4948043<UP,BROADCAST,RUNNING,MULTICAST,TCPCKSUM,NOWINS> mtu 1500 inet 172.1.0.1 netmask 0xffff0000 broadcast 172.1.255.255 partner 172.1.0.2 (not in use) ether 02:a0:98:08:22:b6 (Enabled virtual interface) nfo enabled array02>ifconfig -a e0a: flags=948043<UP,BROADCAST,RUNNING,MULTICAST,TCPCKSUM> mtu 1500 ether 02:a0:98:0c:1a:37 (auto-1000t-fd-up) flowcontrol full trunked lan0 e0b: flags=108042<BROADCAST,RUNNING,MULTICAST,TCPCKSUM> mtu 1500 ether 00:a0:98:0c:1a:36 (auto-unknown-cfg_down) flowcontrol full e0c: flags=948043<UP,BROADCAST,RUNNING,MULTICAST,TCPCKSUM> mtu 1500 ether 02:a0:98:0c:1a:37 (auto-1000t-fd-up) flowcontrol full trunked lan0 e0d: flags=108042<BROADCAST,RUNNING,MULTICAST,TCPCKSUM> mtu 1500 ether 00:a0:98:0c:1a:34 (auto-unknown-cfg_down) flowcontrol full e2a: flags=948043<UP,BROADCAST,RUNNING,MULTICAST,TCPCKSUM> mtu 1500 ether 02:a0:98:0c:1a:36 (auto-10g_sr-fd-up) flowcontrol full trunked tgif e2b: flags=948043<UP,BROADCAST,RUNNING,MULTICAST,TCPCKSUM> mtu 1500 ether 02:a0:98:0c:1a:36 (auto-10g_sr-fd-up) flowcontrol full trunked tgif lo: flags=1948049<UP,LOOPBACK,RUNNING,MULTICAST,TCPCKSUM> mtu 8160 inet 127.0.0.1 netmask 0xff000000 broadcast 127.0.0.1 ether 00:00:00:00:00:00 (VIA Provider) lan0: flags=948043<UP,BROADCAST,RUNNING,MULTICAST,TCPCKSUM> mtu 1500 inet 10.28.17.214 netmask 0xffffff00 broadcast 10.28.17.255 partner 10.28.17.213 (not in use) ether 02:a0:98:0c:1a:37 (Enabled virtual interface) tgif: flags=4948043<UP,BROADCAST,RUNNING,MULTICAST,TCPCKSUM,NOWINS> mtu 1500 inet 172.1.0.2 netmask 0xffff0000 broadcast 172.1.255.255 partner 172.1.0.1 (not in use) ether 02:a0:98:0c:1a:36 (Enabled virtual interface) nfo enabled This message (including any attachments) contains confidential and/or proprietary information intended only for the addressee. Any unauthorized disclosure, copying, distribution or reliance on the contents of this information is strictly prohibited and may constitute a violation of law. If you are not the intended recipient, please notify the sender immediately by responding to this e-mail, and delete the message from your system. If you have any questions about this e-mail please notify the sender immediately. This message (including any attachments) contains confidential and/or proprietary information intended only for the addressee. Any unauthorized disclosure, copying, distribution or reliance on the contents of this information is strictly prohibited and may constitute a violation of law. If you are not the intended recipient, please notify the sender immediately by responding to this e-mail, and delete the message from your system. If you have any questions about this e-mail please notify the sender immediately.

2 3

Re: Multistore Setup
by Villabroza, Gerald 14 May '08

14 May '08

Cristian Rojas wrote: > Hi Gerald, thanks for your questions, these is the output: > > lgsstpnas02> vfiler status -r > vfiler0 running > ipspace: default-ipspace > IP address: 10.223.59.111 [vifp32] > IP address: 192.168.56.4 [vifpriv] > IP address: 192.168.215.95 [e0] > Path: / [/etc] > UUID: 00000000-0000-0000-0000-000000000000 > > excvfiler1 running > ipspace: excfiler-ipspace > IP address: 10.223.61.91 [e6b] > Path: /vol/excvfiler [/etc] > UUID: bd4cee12-1905-11dd-84ee-00a0980159d1 > > > ---------------------------0------------------------- >> 2. is the ip address in the output pingable from someplace? > No, not even from the filer :( It won't be pingable from the filer because routed is off. This is requried for Multistore. Try configuring a system on the same network using an IP in the same subnet. Ping the the vfiler from there and ping the system from the vfiler. > ---------------------------0------------------------- >> 3. are you using vlans or ipspaces? > Ipspaces > > lgsstpnas02> ipspace list > Number of ipspaces configured: 2 > default-ipspace (e0 e9b e5b e7b vifp32 vifpriv) > excfiler-ipspace (e6b) > > ---------------------------0------------------------- >> 4. what does 'vfiler run excvfiler1 run route -s'say? > > lgsstpnas02> vfiler run excvfiler1 route -s > > ===== excvfiler1 > Routing tables > > Internet: > Destination Gateway Flags Refs Use Interface > default 10.223.61.254 UGS 0 0 e6b > 10.223.61/24 link#7 UC 0 0 e6b > 10.223.61.254 link#7 UHL 1 0 e6b That default route doesn't look right for the e6b ip. What subnet mask is used for the e6b ifconfig statement? -=-=- gerald villabroza <geraldv at stanford.edu> technical lead, its storage, stanford university

1 0

RE: cannot create incremental snapshot
by Somavarapu, Nagender 12 May '08

12 May '08

Hi Chaim, setting fractional_reserve to 0 has some side effects with LUNs which could be undesirable. You may like to understand this option in more detail before disabling it. thanks S Nagender -----Original Message----- From: Chaim Rieger [mailto:lists@up-south.com] Sent: Friday, May 09, 2008 12:04 AM To: Stephen C. Losen Cc: toasters(a)mathworks.com Subject: Re: cannot create incremental snapshot Stephen C. Losen wrote: > > You probably need to set the fractional reserve on the source volume > to zero: > > vol options qryndbcore fractional_reserve 0 > > > > Steve Losen scl(a)virginia.edu phone: 434-924-0640 > > University of Virginia ITC Unix Support > that did the trick thanx

1 0

RE: Exports file
by Kevin Parker 09 May '08

09 May '08

Most permissive permission always applies. So here, client will get RW from the "rw=-123.123.4.0/24" statement. ro will be dropped as client gets most permissive permission when conflict. Try it! You can also play with 'exportfs -c' (I think "-c" is the arg - give it a client name and it will show you what exported and how) Best regards, ~~~~~~~~~~~~~~~~ Kevin Parker Mobile: 919.606.8737 http://theparkerz.com <blocked::http://theparkerz.com/> ~~~~~~~~~~~~~~~~ _____ From: owner-toasters(a)mathworks.com [mailto:owner-toasters@mathworks.com] On Behalf Of Conner, Neil Sent: Thursday, May 08, 2008 6:28 PM To: Toasters Subject: Exports file I want configure an exports file to grant a client ro access when I've excluded the subnet the client belongs to from rw access. The man page reads: "You cannot exclude an NFS client identifier from the ro= or rw= option and include the same NFS client identifier in the other option", but I'm wondering if that restriction applies to subnets. This is what I have now: /vol/vol0/dir1 -sec=sys,rw=-123.123.4.0/24:123.123.0.0/16,root=adminhost And I want something like this: /vol/vol0/dir1 -sec=sys,ro=123.123.4.4,rw=-123.123.4.0/24:123.123.0.0/16,root=adminhost Is this possible?

6 8

Re: Max of 20MB/sec through FC attached LTO tape drive? Overland neo 2000
by Holloway, Chris 09 May '08

09 May '08

I have seen the minra option on the volume have a major affect on backup speed. Have you got it set to minra=on on any of your volumes? Cheers Chris ----- Original Message ----- From: owner-toasters(a)mathworks.com <owner-toasters(a)mathworks.com> To: Toasters <toasters(a)mathworks.com> Sent: Fri May 09 16:02:14 2008 Subject: Max of 20MB/sec through FC attached LTO tape drive? Overland neo 2000 Asking here because we don't have current Overland support for the fibre channel card and I would guess it is more of a problem with the card rather than netapp (we only used the FCO with a netapp), but maybe someone has encountered this... Around 2004 we bought a Overland Neo 2000 tape library with the FCO VIA scsi to FC converter board, and we got at most 20 or 25MB/sec through it (per tape drive) depending if we were writing or reading. This is both with NDMP dumps from the filer disks to tape, and also across the network into the filer to tape with NDMP. The backup software is Bakbone Netvault. Last year we tried attaching this library to a Unix host directly with SCSI and our speed almost doubled. I was floored, and shocked that we spent so many years missing out on performance, although 20-25 was the best I've seen anyone report for LTO2 so I never expected more. Recently we bought another Neo 2000 with LTO4 drives, and I decided to move the FCO card over to it and try again, hoping maybe there was a misconfiguration, but I could not find any and it still acts as if there is a mystery cap on the speed of 20MB/sec (it averages 18-19 after it gets going). I've been all through the configuration of the FCO card and it is set to 160M/sec per scsi bus, tried with and without domain validation, and I even got the FC channel to operate at 2Gbit but no improvement. If I cannot figure this out then I will just attach it to the Unix host with the other neo 2000, but its always nice to have more options. I would not bother operating a LTO4 drive at such a slow speed in production because shoe-shining would probably wear it out and the compression would be sub optimal or non-existent. -------------------------------------------------------- Christopher Holloway Lead Consultant Professional Services Main: +44 (0)1423 519000 Direct: +44 (0)1423 519099 Mobile: +44 (0)7867 783584 Fax: +44 (0)1423 504415 DNS Arrow launches new enhanced customer portal - Inform an e-commerce partner tool to help customers save time and money in procurement. For more information on Inform, please contact your account manager on 0800 983 2525 or register for an account at www.dnsarrow.co.uk/inform DNS Arrow Nidderdale House Beckwith Knowle Harrogate HG3 1SA www.DNSarrow.co.uk Email Confidentiality Notice: This message is private and confidential. If you have received this message in error, please notify us and remove it from your system. "DNS Arrow" means Digital Network Services (UK) Limited, a company registered in England. Company Registration Number 3952678. Registered Office: Nidderdale House, Otley Road, Beckwith Knowle, Harrogate HG3 1SA, UK.

2 1

Re: Max of 20MB/sec through FC attached LTO tape drive? Overland neo 2000
by Webster, Stetson 09 May '08

09 May '08

Hey Folks, Bringing up 'minra' starts to tread into a performance issue. I highly recommend that a NetApp support case be opened to ensure that this is handled with the appropriate skillset, experience and attention. Pardon brevity, typos, etc; sent from mobile. -----Original Message----- From: Holloway, Chris [mailto:chris.holloway@DNSArrow.co.uk] Sent: Friday, May 09, 2008 12:26 PM Pacific Standard Time To: mcdouga9(a)egr.msu.edu; toasters(a)mathworks.com Subject: Re: Max of 20MB/sec through FC attached LTO tape drive? Overland neo 2000 I have seen the minra option on the volume have a major affect on backup speed. Have you got it set to minra=on on any of your volumes? Cheers Chris ----- Original Message ----- From: owner-toasters(a)mathworks.com <owner-toasters(a)mathworks.com> To: Toasters <toasters(a)mathworks.com> Sent: Fri May 09 16:02:14 2008 Subject: Max of 20MB/sec through FC attached LTO tape drive? Overland neo 2000 Asking here because we don't have current Overland support for the fibre channel card and I would guess it is more of a problem with the card rather than netapp (we only used the FCO with a netapp), but maybe someone has encountered this... Around 2004 we bought a Overland Neo 2000 tape library with the FCO VIA scsi to FC converter board, and we got at most 20 or 25MB/sec through it (per tape drive) depending if we were writing or reading. This is both with NDMP dumps from the filer disks to tape, and also across the network into the filer to tape with NDMP. The backup software is Bakbone Netvault. Last year we tried attaching this library to a Unix host directly with SCSI and our speed almost doubled. I was floored, and shocked that we spent so many years missing out on performance, although 20-25 was the best I've seen anyone report for LTO2 so I never expected more. Recently we bought another Neo 2000 with LTO4 drives, and I decided to move the FCO card over to it and try again, hoping maybe there was a misconfiguration, but I could not find any and it still acts as if there is a mystery cap on the speed of 20MB/sec (it averages 18-19 after it gets going). I've been all through the configuration of the FCO card and it is set to 160M/sec per scsi bus, tried with and without domain validation, and I even got the FC channel to operate at 2Gbit but no improvement. If I cannot figure this out then I will just attach it to the Unix host with the other neo 2000, but its always nice to have more options. I would not bother operating a LTO4 drive at such a slow speed in production because shoe-shining would probably wear it out and the compression would be sub optimal or non-existent. -------------------------------------------------------- Christopher Holloway Lead Consultant Professional Services Main: +44 (0)1423 519000 Direct: +44 (0)1423 519099 Mobile: +44 (0)7867 783584 Fax: +44 (0)1423 504415 DNS Arrow launches new enhanced customer portal - Inform an e-commerce partner tool to help customers save time and money in procurement. For more information on Inform, please contact your account manager on 0800 983 2525 or register for an account at www.dnsarrow.co.uk/inform DNS Arrow Nidderdale House Beckwith Knowle Harrogate HG3 1SA www.DNSarrow.co.uk Email Confidentiality Notice: This message is private and confidential. If you have received this message in error, please notify us and remove it from your system. "DNS Arrow" means Digital Network Services (UK) Limited, a company registered in England. Company Registration Number 3952678. Registered Office: Nidderdale House, Otley Road, Beckwith Knowle, Harrogate HG3 1SA, UK.

1 0

Max of 20MB/sec through FC attached LTO tape drive? Overland neo 2000
by Adam McDougall 09 May '08

09 May '08

Asking here because we don't have current Overland support for the fibre channel card and I would guess it is more of a problem with the card rather than netapp (we only used the FCO with a netapp), but maybe someone has encountered this... Around 2004 we bought a Overland Neo 2000 tape library with the FCO VIA scsi to FC converter board, and we got at most 20 or 25MB/sec through it (per tape drive) depending if we were writing or reading. This is both with NDMP dumps from the filer disks to tape, and also across the network into the filer to tape with NDMP. The backup software is Bakbone Netvault. Last year we tried attaching this library to a Unix host directly with SCSI and our speed almost doubled. I was floored, and shocked that we spent so many years missing out on performance, although 20-25 was the best I've seen anyone report for LTO2 so I never expected more. Recently we bought another Neo 2000 with LTO4 drives, and I decided to move the FCO card over to it and try again, hoping maybe there was a misconfiguration, but I could not find any and it still acts as if there is a mystery cap on the speed of 20MB/sec (it averages 18-19 after it gets going). I've been all through the configuration of the FCO card and it is set to 160M/sec per scsi bus, tried with and without domain validation, and I even got the FC channel to operate at 2Gbit but no improvement. If I cannot figure this out then I will just attach it to the Unix host with the other neo 2000, but its always nice to have more options. I would not bother operating a LTO4 drive at such a slow speed in production because shoe-shining would probably wear it out and the compression would be sub optimal or non-existent.

2 2

Exports file
by Conner, Neil 08 May '08

08 May '08

I want configure an exports file to grant a client ro access when I¹ve excluded the subnet the client belongs to from rw access. The man page reads: ³You cannot exclude an NFS client identifier from the ro= or rw= option and include the same NFS client identifier in the other option², but I¹m wondering if that restriction applies to subnets. This is what I have now: /vol/vol0/dir1 -sec=sys,rw=-123.123.4.0/24:123.123.0.0/16,root=adminhost And I want something like this: /vol/vol0/dir1 -sec=sys,ro=123.123.4.4,rw=-123.123.4.0/24:123.123.0.0/16,root=adminhost Is this possible?

1 0

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 May 2008