toasters

toasters@lists.teaparty.net

13515 discussions

Re: dlt autoloaders
by Darrell Fuhriman 09 Sep '00

09 Sep '00

Brian Tao <taob(a)risc.org> writes: > Out of curiosity, what's the aggregate throughput on your backups > to 12 DLT7000's? They should be able to do, what, 5MB/sec if you can > stream data to them consistently? 12 drives would be 60MB/sec... are You can push much more than that -- we sustain 10-12MB/sec to a DLT7000. > you actually seeing that much throughput over the network? Yes, we see 60MB/sec, and have pushed more in tests. Darrell

1 0

Re: gigabit ethernet performance
by Eyal Traitel 09 Sep '00

09 Sep '00

Only on 6.0 buddy... Eyal. --- Brian Tao <taob(a)risc.org> wrote: > On Thu, 7 Sep 2000, Eyal Traitel wrote: > > > > Have you considered Jumbo packets ? It could help you... > > How do you set a jumbo frame size on a Netapp? I recall trying a > few combinations of ifconfig invocations, but nothing would bump it > up > to 9K. > -- > Brian Tao (BT300, taob(a)risc.org) > "Though this be madness, yet there is method in't" > ===== Yours, Eyal Traitel eTraitel(a)yahoo.com, Home: 972-3-5290415 (Tel Aviv) *** eTraitel - it's the new eBuzzword ! *** __________________________________________________ Do You Yahoo!? Yahoo! Mail - Free email you can access from anywhere! http://mail.yahoo.com/

1 0

Re: gigabit ethernet performance
by Eyal Traitel 09 Sep '00

09 Sep '00

Marcus, Have you considered Jumbo packets ? It could help you... Eyal. Marcus Smiley <marcus.smiley(a)ericsson.com> wrote: Does anyone have any benchmarks/suggestions for poor performace with gigabit ethernet between Sun enterprise 220R and NetApp 740 with Intel GBII card? I have an open ticket w/Netapp and it's been open forever. These machines are directly attached (private network), and I've tried a mkfile and copying that file back during the day and at night (no load) and it makes no difference. The performance is ~ 10 MB/s, no matter what. The Sun is running Solaris 8, with latest GB Ethernet 3.0 drivers and patches. Also, the Sun box's only function is Legato server, so I've even stopped all nsr daemons while the test copy/mkfile is in progress, & it makes no difference. I've tried NFSv2 vs. NFSv3, with 32k size packets and 8k size packets. I've tried tcp and udp, with the same results. The /etc/system file on the E220R looks like this, for the ge adapter: * Increase size of STREAMS syncronized queues to increase network * performance. Release notes for qfe card recommends setting this * to 25 per 64Mbytes of RAM in the system. It also prevents * receive overrun on the GbE interface. msmiley 7/20/00 set sq_max_size=775 set ge:ge_adv_1000autoneg_cap=0 set ge:ge_adv_1000fdx_cap=1 set ge:ge_adv_pauseTX=1 Also, I've set tcp/udp specific stuff per the Netapp tech suppt. folks: ndd -set /dev/udp udp_recv_hiwat 65535 ndd -set /dev/udp udp_xmit_hiwat 65535 ndd -set /dev/tcp tcp_recv_hiwat 65535 ndd -set /dev/tcp tcp_xmit_hiwat 65535 The netapp is a 740 (5.3.6R1P1) with over 100GB of free space available on the volume I'm copying to/from. Interestingly enough, I have done a packet trace (pktt) on the GBII card and it looks OK. Also, the throughput from my 100BaseT card to the netapp (thru a catalyst 5500) does ~ 7.5 MB/s. *Any* input here would be much appreciated. -Marc Marcus Smiley | Ericsson Wireless Communications, Inc. Unix Sys Admin | 6210 Spine Rd. Boulder, CO 80301 | Ph (303) 473-6903 Yours, Eyal Traitel eTraitel(a)yahoo.com, Home: 972-3-5290415 (Tel Aviv) *** eTraitel - it's the new eBuzzword ! *** --------------------------------- Do You Yahoo!? Yahoo! Mail - Free email you can access from anywhere!

2 1

gigabit ethernet performance
by Marcus Smiley 09 Sep '00

09 Sep '00

Does anyone have any benchmarks/suggestions for poor performace with gigabit ethernet between Sun enterprise 220R and NetApp 740 with Intel GBII card? I have an open ticket w/Netapp and it's been open forever. These machines are directly attached (private network), and I've tried a mkfile and copying that file back during the day and at night (no load) and it makes no difference. The performance is ~ 10 MB/s, no matter what. The Sun is running Solaris 8, with latest GB Ethernet 3.0 drivers and patches. Also, the Sun box's only function is Legato server, so I've even stopped all nsr daemons while the test copy/mkfile is in progress, & it makes no difference. I've tried NFSv2 vs. NFSv3, with 32k size packets and 8k size packets. I've tried tcp and udp, with the same results. The /etc/system file on the E220R looks like this, for the ge adapter: * Increase size of STREAMS syncronized queues to increase network * performance. Release notes for qfe card recommends setting this * to 25 per 64Mbytes of RAM in the system. It also prevents * receive overrun on the GbE interface. msmiley 7/20/00 set sq_max_size=775 set ge:ge_adv_1000autoneg_cap=0 set ge:ge_adv_1000fdx_cap=1 set ge:ge_adv_pauseTX=1 Also, I've set tcp/udp specific stuff per the Netapp tech suppt. folks: ndd -set /dev/udp udp_recv_hiwat 65535 ndd -set /dev/udp udp_xmit_hiwat 65535 ndd -set /dev/tcp tcp_recv_hiwat 65535 ndd -set /dev/tcp tcp_xmit_hiwat 65535 The netapp is a 740 (5.3.6R1P1) with over 100GB of free space available on the volume I'm copying to/from. Interestingly enough, I have done a packet trace (pktt) on the GBII card and it looks OK. Also, the throughput from my 100BaseT card to the netapp (thru a catalyst 5500) does ~ 7.5 MB/s. *Any* input here would be much appreciated. -Marc Marcus Smiley | Ericsson Wireless Communications, Inc. Unix Sys Admin | 6210 Spine Rd. Boulder, CO 80301 | Ph (303) 473-6903

2 1

dlt autoloaders
by neil lehrer 09 Sep '00

09 Sep '00

hi, i'd like recommendations for a dlt autoloader to be attached to an f760 and controlled by legato. thanks. --

5 6

RE: Deleting snapshot
by Premanshu Jain 09 Sep '00

09 Sep '00

Ndmpd status, doesn't revealany active ndmp sessions. These snapshots date back to 17 Aug. Any other way out..avoiding reboot prem -----Original Message----- From: Steve Kappel [mailto:steve.kappel@raistlin.min.ov.com] Sent: Friday, September 08, 2000 1:22 PM To: mrb(a)netapp.com; PrJain(a)shastanets.com Cc: toasters(a)mathworks.com Subject: RE: Deleting snapshot No, not a chance that it is the backup software (NDMP client). Consider this - if your network link goes down your backup (NDMP client) is aborted. Since you have no network link there is no way the backup product can do anything in NDMP to abort or cleanup the backup running in the filer. The filer has to be able to clean up from the NDMP session socket "going away". Since you have "failed backup processes" the NDMP session is closed on the client side. So it is OnTap that is not functioning correctly. You might try looking for the NDMP sessions with "ndmpd status" on the filer and then kill any that are running with "ndmpd kill <sessionid>". >From: Premanshu Jain <PrJain(a)shastanets.com> >To: "'mrb(a)netapp.com'" <mrb(a)netapp.com>, Premanshu Jain <PrJain(a)shastanets.com> >Cc: toasters(a)mathworks.com >Subject: RE: Deleting snapshot >Date: Fri, 8 Sep 2000 11:09:01 -0700 > >NDMP or Veritas Netbackup. All these snapshots were a part of failed backups >prem > >-----Original Message----- >From: J. Bryan Kelsch [mailto:mrb@netapp.com] >Sent: Friday, September 08, 2000 11:13 AM >To: Premanshu Jain >Cc: toasters(a)mathworks.com >Subject: Re: Deleting snapshot > > >This is not a function the filer, it is an error in the backup software that >is not cleaning up properly. > > > > >> >> >> >>You have to reboot first. >> >> >>On Fri, 8 Sep 2000 10:42:26 -0700 , you wrote: >> >>>I have some snapshots created by failed backup processes and shows them >busy >>>still. How can I delete them..? >>> >>> >>>Volume sw >>>working.................................................................. . >.. >>>......................................................................... . >.. >>>.................................... >>> >>> %/used %/total date name >>>---------- ---------- ------------ -------- >>> 3% ( 3%) 2% ( 2%) Sep 07 18:04 bot_sw.1 >>> 4% ( 3%) 3% ( 2%) Sep 07 16:12 hourly.0 >>> 4% ( 4%) 3% ( 3%) Sep 07 12:10 hourly.1 >>> 5% ( 4%) 3% ( 3%) Sep 07 08:10 hourly.2 >>> 5% ( 4%) 4% ( 3%) Sep 07 00:11 nightly.0 >>> 6% ( 4%) 4% ( 3%) Sep 06 20:11 hourly.3 >>> 7% ( 4%) 5% ( 3%) Sep 06 16:12 hourly.4 >>> 7% ( 4%) 5% ( 3%) Sep 06 12:10 hourly.5 >>> 8% ( 4%) 6% ( 3%) Sep 06 00:08 nightly.1 >>> 19% (15%) 17% (12%) Aug 20 21:52 snapshot_for_backup.32 (busy) >>> 20% (15%) 17% (12%) Aug 19 22:46 snapshot_for_backup.30 (busy) >>> 21% (15%) 19% (12%) Aug 17 04:08 snapshot_for_backup.23 (busy) > > >------------------------ >J. Bryan Kelsch, PSE >Network Appliance >5220 Spring Valley Road - Suite 350 >Dallas, TX 75240 >972.759.7528 Office >817.821.0350 Cell >888.374.6422 Pager __________________________________________________________________________ Steve Kappel steve.kappel(a)veritas.com VERITAS Software steve.kappel(a)iname.com (Personal)

3 2

multiple raid grps in filesystem on filer
by neil lehrer 08 Sep '00

08 Sep '00

hi, i understand that having more than 1 raid grp in a filesystem makes the filer more fault tolerant of a full raid grp failure. so lets say there are two raid grps in the filesystem. if you lose a raid grp [2 drives fail] how do you restore it? i can understand restoring the whole filesystem or a qtree or a dir, but if files and dirs can go across more than 1 raid grp in a filesystem what is the advantage of short restore of the broken raid grp? how do you put back just that raid group? i know you can replace the bad drives and then recreate the group but how do you restore the data for just that raid grp? thanks

2 1

RE: Deleting snapshot
by Steve Kappel 08 Sep '00

08 Sep '00

No, not a chance that it is the backup software (NDMP client). Consider this - if your network link goes down your backup (NDMP client) is aborted. Since you have no network link there is no way the backup product can do anything in NDMP to abort or cleanup the backup running in the filer. The filer has to be able to clean up from the NDMP session socket "going away". Since you have "failed backup processes" the NDMP session is closed on the client side. So it is OnTap that is not functioning correctly. You might try looking for the NDMP sessions with "ndmpd status" on the filer and then kill any that are running with "ndmpd kill <sessionid>". >From: Premanshu Jain <PrJain(a)shastanets.com> >To: "'mrb(a)netapp.com'" <mrb(a)netapp.com>, Premanshu Jain <PrJain(a)shastanets.com> >Cc: toasters(a)mathworks.com >Subject: RE: Deleting snapshot >Date: Fri, 8 Sep 2000 11:09:01 -0700 > >NDMP or Veritas Netbackup. All these snapshots were a part of failed backups >prem > >-----Original Message----- >From: J. Bryan Kelsch [mailto:mrb@netapp.com] >Sent: Friday, September 08, 2000 11:13 AM >To: Premanshu Jain >Cc: toasters(a)mathworks.com >Subject: Re: Deleting snapshot > > >This is not a function the filer, it is an error in the backup software that >is not cleaning up properly. > > > > >> >> >> >>You have to reboot first. >> >> >>On Fri, 8 Sep 2000 10:42:26 -0700 , you wrote: >> >>>I have some snapshots created by failed backup processes and shows them >busy >>>still. How can I delete them..? >>> >>> >>>Volume sw >>>working................................................................... >.. >>>.......................................................................... >.. >>>.................................... >>> >>> %/used %/total date name >>>---------- ---------- ------------ -------- >>> 3% ( 3%) 2% ( 2%) Sep 07 18:04 bot_sw.1 >>> 4% ( 3%) 3% ( 2%) Sep 07 16:12 hourly.0 >>> 4% ( 4%) 3% ( 3%) Sep 07 12:10 hourly.1 >>> 5% ( 4%) 3% ( 3%) Sep 07 08:10 hourly.2 >>> 5% ( 4%) 4% ( 3%) Sep 07 00:11 nightly.0 >>> 6% ( 4%) 4% ( 3%) Sep 06 20:11 hourly.3 >>> 7% ( 4%) 5% ( 3%) Sep 06 16:12 hourly.4 >>> 7% ( 4%) 5% ( 3%) Sep 06 12:10 hourly.5 >>> 8% ( 4%) 6% ( 3%) Sep 06 00:08 nightly.1 >>> 19% (15%) 17% (12%) Aug 20 21:52 snapshot_for_backup.32 (busy) >>> 20% (15%) 17% (12%) Aug 19 22:46 snapshot_for_backup.30 (busy) >>> 21% (15%) 19% (12%) Aug 17 04:08 snapshot_for_backup.23 (busy) > > >------------------------ >J. Bryan Kelsch, PSE >Network Appliance >5220 Spring Valley Road - Suite 350 >Dallas, TX 75240 >972.759.7528 Office >817.821.0350 Cell >888.374.6422 Pager __________________________________________________________________________ Steve Kappel steve.kappel(a)veritas.com VERITAS Software steve.kappel(a)iname.com (Personal)

2 1

RE: Deleting snapshot
by Premanshu Jain 08 Sep '00

08 Sep '00

Oh..This is CRAZY on part of NetAPPS...!!!!!!!!!! -----Original Message----- From: J. Bryan Kelsch [mailto:mrb@netapp.com] Sent: Friday, September 08, 2000 10:56 AM To: Premanshu Jain Subject: Re: Deleting snapshot You have to reboot first. On Fri, 8 Sep 2000 10:42:26 -0700 , you wrote: >I have some snapshots created by failed backup processes and shows them busy >still. How can I delete them..? > > >Volume sw >working.................................................................... . >........................................................................... . >.................................... > > %/used %/total date name >---------- ---------- ------------ -------- > 3% ( 3%) 2% ( 2%) Sep 07 18:04 bot_sw.1 > 4% ( 3%) 3% ( 2%) Sep 07 16:12 hourly.0 > 4% ( 4%) 3% ( 3%) Sep 07 12:10 hourly.1 > 5% ( 4%) 3% ( 3%) Sep 07 08:10 hourly.2 > 5% ( 4%) 4% ( 3%) Sep 07 00:11 nightly.0 > 6% ( 4%) 4% ( 3%) Sep 06 20:11 hourly.3 > 7% ( 4%) 5% ( 3%) Sep 06 16:12 hourly.4 > 7% ( 4%) 5% ( 3%) Sep 06 12:10 hourly.5 > 8% ( 4%) 6% ( 3%) Sep 06 00:08 nightly.1 > 19% (15%) 17% (12%) Aug 20 21:52 snapshot_for_backup.32 (busy) > 20% (15%) 17% (12%) Aug 19 22:46 snapshot_for_backup.30 (busy) > 21% (15%) 19% (12%) Aug 17 04:08 snapshot_for_backup.23 (busy) ------------------------ J. Bryan Kelsch, PSE Network Appliance 5220 Spring Valley Road - Suite 350 Dallas, TX 75240 972.759.7528 Office 817.821.0350 Cell 888.374.6422 Pager

2 1

Deleting snapshot
by Premanshu Jain 08 Sep '00

08 Sep '00

I have some snapshots created by failed backup processes and shows them busy still. How can I delete them..? Volume sw working..................................................................... ............................................................................ .................................... %/used %/total date name ---------- ---------- ------------ -------- 3% ( 3%) 2% ( 2%) Sep 07 18:04 bot_sw.1 4% ( 3%) 3% ( 2%) Sep 07 16:12 hourly.0 4% ( 4%) 3% ( 3%) Sep 07 12:10 hourly.1 5% ( 4%) 3% ( 3%) Sep 07 08:10 hourly.2 5% ( 4%) 4% ( 3%) Sep 07 00:11 nightly.0 6% ( 4%) 4% ( 3%) Sep 06 20:11 hourly.3 7% ( 4%) 5% ( 3%) Sep 06 16:12 hourly.4 7% ( 4%) 5% ( 3%) Sep 06 12:10 hourly.5 8% ( 4%) 6% ( 3%) Sep 06 00:08 nightly.1 19% (15%) 17% (12%) Aug 20 21:52 snapshot_for_backup.32 (busy) 20% (15%) 17% (12%) Aug 19 22:46 snapshot_for_backup.30 (busy) 21% (15%) 19% (12%) Aug 17 04:08 snapshot_for_backup.23 (busy)

2 1

RE: Deleting snapshot
by Premanshu Jain 08 Sep '00

08 Sep '00

NDMP or Veritas Netbackup. All these snapshots were a part of failed backups prem -----Original Message----- From: J. Bryan Kelsch [mailto:mrb@netapp.com] Sent: Friday, September 08, 2000 11:13 AM To: Premanshu Jain Cc: toasters(a)mathworks.com Subject: Re: Deleting snapshot This is not a function the filer, it is an error in the backup software that is not cleaning up properly. > > > >You have to reboot first. > > >On Fri, 8 Sep 2000 10:42:26 -0700 , you wrote: > >>I have some snapshots created by failed backup processes and shows them busy >>still. How can I delete them..? >> >> >>Volume sw >>working................................................................... .. >>.......................................................................... .. >>.................................... >> >> %/used %/total date name >>---------- ---------- ------------ -------- >> 3% ( 3%) 2% ( 2%) Sep 07 18:04 bot_sw.1 >> 4% ( 3%) 3% ( 2%) Sep 07 16:12 hourly.0 >> 4% ( 4%) 3% ( 3%) Sep 07 12:10 hourly.1 >> 5% ( 4%) 3% ( 3%) Sep 07 08:10 hourly.2 >> 5% ( 4%) 4% ( 3%) Sep 07 00:11 nightly.0 >> 6% ( 4%) 4% ( 3%) Sep 06 20:11 hourly.3 >> 7% ( 4%) 5% ( 3%) Sep 06 16:12 hourly.4 >> 7% ( 4%) 5% ( 3%) Sep 06 12:10 hourly.5 >> 8% ( 4%) 6% ( 3%) Sep 06 00:08 nightly.1 >> 19% (15%) 17% (12%) Aug 20 21:52 snapshot_for_backup.32 (busy) >> 20% (15%) 17% (12%) Aug 19 22:46 snapshot_for_backup.30 (busy) >> 21% (15%) 19% (12%) Aug 17 04:08 snapshot_for_backup.23 (busy) ------------------------ J. Bryan Kelsch, PSE Network Appliance 5220 Spring Valley Road - Suite 350 Dallas, TX 75240 972.759.7528 Office 817.821.0350 Cell 888.374.6422 Pager

1 0

RE: vif between 740 & 520 problems & questions
by Thompson, Arnold 08 Sep '00

08 Sep '00

Hi John, You're not doing anything wrong - it's the load-balancing algorithm used by etherchannel that's causing you to use only one interface. In a nutshell, etherchannel does a hash on the last two bits of the source MAC address and comes up with one of four possible combinations. Your NDMPcopy connection starts up on only one port, using a single MAC address, so essentially there are no multiple MAC addresses to load-balance. Additionally, the balancing decision based on this hash is made by an etherchannel-enabled switch, not the Netapp box itself. Etherchannel requires all ports to be on the same switch and only really works when you have multiple clients (and consequently multiple possible hash combinations) accessing one server - it fails completely when you have a single "client" accessing a single server. In brief, to get the best migration performance, use gigabit ethernet between two systems. You can simply reverse the optical cable leads - you don't even need a switch. The good news is that more sophisticated load-balancing algorithms are under consideration, although I can't give you specifics or tell you when Netapp will have them available. Good luck, Arnie > -----Original Message----- > From: John Stoffel [mailto:john@casc.com] > Sent: Friday, September 08, 2000 7:11 AM > To: toasters(a)mathworks.com > Subject: vif between 740 & 520 problems & questions > > > > Hi all, > > I've got a couple of questions on setting up and using vifs between > two filers to help speed up data migration from the old F520 to a new > F740. > > They're configed like this: > > F520 F740 > OnTap: 5.2.1 OnTap: 5.3.6p2 > RAM: 256Mb RAM: 512Mb > Quad Ethernet Quad Ethernet > Gigabit Ethernet > > I've networked the two systems together using three ports on each Quad > Ethernet board, but due to distance limitations, I'm going through > three seperate hubs. So the two filers are basically hooked up > back-to-back over a private network. > > I did the various vif commands, and they are setup correctly, and they > are passing traffic properly, but just over one of the three links. I > know I need to upgrade the F520 to match OnTap versions, and will do > so next weekend when I have some down time, but I wanted to play > around with ndmpcopy beforehand to see how the performance will be. > > Doing a test this morning, I managed to crash the new F740 with a PCI > Parity error while doing another test of ndmpcopy. Whee! > > My questions: > > - will vifs work properly across direct connections between filers? > > - Has anyone used a Lucent Cajun P220 switch to implement a VIF? I > could easily run a fiber link from the F740 to a Cajun which would > then run the three ethernet links. > > And while I work for Lucent, the Cajun wasn't made in our division and > I haven't used it. > > Thanks, > John > John Stoffel - Senior Unix Systems Administrator - Lucent > Technologies > stoffel(a)lucent.com - http://www.lucent.com - 978-952-7548 > john.stoffel(a)ascend.com - http://www.ascend.com >

1 0

SUMMARY: vif between 740 & 520 problems & questions
by John Stoffel 08 Sep '00

08 Sep '00

Hi all, The general consensus is that a VIF won't help me at all, since all the data will just go via the primary interface, and won't use the two other interfaces at all. This is what I'm seeing on the system anyway, so I guess I'm out of luck. Thanks for all your comments! John John Stoffel - Senior Unix Systems Administrator - Lucent Technologies stoffel(a)lucent.com - http://www.lucent.com - 978-952-7548 john.stoffel(a)ascend.com - http://www.ascend.com

1 0

vif between 740 & 520 problems & questions
by John Stoffel 08 Sep '00

08 Sep '00

Hi all, I've got a couple of questions on setting up and using vifs between two filers to help speed up data migration from the old F520 to a new F740. They're configed like this: F520 F740 OnTap: 5.2.1 OnTap: 5.3.6p2 RAM: 256Mb RAM: 512Mb Quad Ethernet Quad Ethernet Gigabit Ethernet I've networked the two systems together using three ports on each Quad Ethernet board, but due to distance limitations, I'm going through three seperate hubs. So the two filers are basically hooked up back-to-back over a private network. I did the various vif commands, and they are setup correctly, and they are passing traffic properly, but just over one of the three links. I know I need to upgrade the F520 to match OnTap versions, and will do so next weekend when I have some down time, but I wanted to play around with ndmpcopy beforehand to see how the performance will be. Doing a test this morning, I managed to crash the new F740 with a PCI Parity error while doing another test of ndmpcopy. Whee! My questions: - will vifs work properly across direct connections between filers? - Has anyone used a Lucent Cajun P220 switch to implement a VIF? I could easily run a fiber link from the F740 to a Cajun which would then run the three ethernet links. And while I work for Lucent, the Cajun wasn't made in our division and I haven't used it. Thanks, John John Stoffel - Senior Unix Systems Administrator - Lucent Technologies stoffel(a)lucent.com - http://www.lucent.com - 978-952-7548 john.stoffel(a)ascend.com - http://www.ascend.com

1 0

FC card slot exchange helped us [Was : gigabit ethernet performance]
by kubek 08 Sep '00

08 Sep '00

I do not have a solution but maybe these experiments done with a 720 may help (I must thank our local reseller support here :). We received an F720 some weeks ago (quad ethernet / FC9) A 'systat 1' showed a really bad performance on the administrative Ethernet (100M) interface : the filer was inputting ~4MB/s when creating a 100MB file from a Sparc. By moving the FC card from the slot 8 to the slot 2 and rebooting the (little:() beast we are now going to ~11MB/s . Seems that netapp hardware layer is as poor as some PC cards builders :( Maybe the trick will work for you : try to move your FC card from its current slot (8?) to another one. Jean-Marie

1 0

Budtool woes
by Clawson, Simon 07 Sep '00

07 Sep '00

Hi! We have Budtool 4.6.1, and are currently looking for a replacement. However... I am bacing up a single volume filer with nearly 200GB of data - my backup is constanlty failing due to the process running out of time..anyone seen this? Is there a way round it? Can I increse the allowable time? Can I split the backup into several levels? HELP! Simon Simon Clawson Renoir Group Systems Administrator Mentor Graphics Uk Rivergate London Road Newbury Berkshire RG14 2QB <<Clawson, Simon.vcf>>

4 4

question1:exportfs -au crash ORACLE daemon; question2:snapshot di r. confusion
by Li, Jack 07 Sep '00

07 Sep '00

Hi, Question#1: We have a F740 filer used as the Oracle DB storage. Our platform is : Sun-Solaris 2.7, SUN 420R,using Giga-bit private connection into filer; using recommended NFS mount options for Oracle 816. Situation: when we have a request to "expand" our filer for sharing (exportfs), and as the time the Oracle daemons were running. We changed the filer's /etc/exports, then executed "exportfs -au", "exportfs -a". And, right after those 2 command to the filer, ALL our Oracle daemons crashed and died. Is that supposed to happen? Question#2: We tried to write a SHELL script to automate our daily (filer) snapshot operation. During the course of testing, we found out a very annoying situation. (The HW/SW specification are the same as above.) There is a Q-tree called "/vol/vol01/ismis_oracle_u1" and is mounted under this machine at the mount point of "/oracle/u1". The problem happened after the snapshot was taken. One of our team member set the source PATH (wrong) to /oracle/u1/oradata/.snapshot in the script (supposed to be "/oracle/u1/.snapshot"). And, this PATH actually worked some of the time. But, when it failed , it showed the error message as "cp: can not find files /oracle/u1/oradata/.snapshot/*". Interesting thing is if we insert a "/bin/ls /orcale/u1/oradata/.snapshot/*" right before the "cp" command, the directory seems to be there after the command , hence the "cp" operation was happy. Any idea? Thanks! Jack Li UNIX System Admin. Team, Toronto Production Support, Capital Markets, Nesbitt Burns Emf!sys, Division of Bank of Montreal Group of Company 416-867-3527,(fax)416-867-7157 <mailto:Jack.Li@bmo.com> Jack.Li(a)bmo.com

3 2

RE: dlt autoloaders
by Premanshu Jain 07 Sep '00

07 Sep '00

I would like to discuss with you in detail about implementing NDMP via gig bit .Let me know the best time I should call. thanks prem Premanshu Jain IS Engineering Support Lead Nortel Networks, Carrier Packet Solutions IP Services Division 2305, Mission College Blvd, Santa Clara, CA 95054 Direct: (408) 565-3573 ESN: 655-3573 eMail: prjain(a)shastanets.com Web: http://www.nortelnetworks.com/ipservices -----Original Message----- From: Kamenzin, Marcus [mailto:kamenzin@netapp.com] Sent: Thursday, September 07, 2000 1:57 PM To: Premanshu Jain; 'Stephen J. Wilcox'; neil lehrer Cc: toasters Subject: RE: dlt autoloaders Hi there, a) AIT II is really good!! b) Quick Restore is the best NDMP-Solution (best NDMP-implementation - really 3-way) c) Don't use FC-Backup, only if you really want the advantage of tape sharing. The supported FC-Backup-solution has a lot of limitations. So, use direct att. SCSI or new Libraries with implemented NDMP-Servers via GBit. Regards Marcus Marcus Kamenzin System Engineer Network Appliance GmbH Unterreut 6 D-76135 Karlsruhe Tel.: (0721) 86901-17 Fax: (0721) 86901-11 Mobile: (0170) 4464341 Mail: kamenzin(a)netapp.com > -----Original Message----- > From: Premanshu Jain [SMTP:PrJain@shastanets.com] > Sent: Thursday, September 07, 2000 6:15 PM > To: 'Stephen J. Wilcox'; neil lehrer > Cc: toasters > Subject: RE: dlt autoloaders > > a) Why don't you try AIT II instead of DLT. I am getting excellant > performance on those 8 mm. Writing more that 100 GB per tape. We are using > Qualstar and spectra logic tape libraries. > > b) Veritas seems to be OK. With lot of limitations and bugs. I have heard of > a better s/w from www.worksta.com > > c) You may use a Vixel Fibre channal switch with fibre channal interfaces on > the concerened servers and the Filers. This will enable accessing tpaes from > anywhere. > > -----Original Message----- > From: Stephen J. Wilcox [mailto:steve@opaltelecom.co.uk] > Sent: Thursday, September 07, 2000 1:07 AM > To: neil lehrer > Cc: toasters > Subject: Re: dlt autoloaders > > > Actually on that subject I am looking to buy a system soon, I'm thinking > of getting an HP surestore library and using Veritas. Theres a lot of > systems to back up altho they dont have that much data on them, then I > have (for now 1) netfiler, does anyone > > a) think surestore are good? > b) think veritas is good? > c) know if i can do anything to save time by directly connecting the tape > library to the filer hence bypassing the server but still keeping the tape > accessible from the server in order to do my network backups? > > Steve > > -- > Stephen J. Wilcox > Internet Manager, Opal Telecom > http://www.opaltelecom.co.uk/ > Tel: 0161 222 2000 > Fax: 0161 222 2008 > > On Wed, 6 Sep 2000, neil lehrer wrote: > > > hi, > > > > i'd like recommendations for a dlt autoloader to be attached to an f760 > > and controlled by legato. thanks. > > -- > >

1 0

RE: dlt autoloaders
by Kamenzin, Marcus 07 Sep '00

07 Sep '00

Hi there, a) AIT II is really good!! b) Quick Restore is the best NDMP-Solution (best NDMP-implementation - really 3-way) c) Don't use FC-Backup, only if you really want the advantage of tape sharing. The supported FC-Backup-solution has a lot of limitations. So, use direct att. SCSI or new Libraries with implemented NDMP-Servers via GBit. Regards Marcus Marcus Kamenzin System Engineer Network Appliance GmbH Unterreut 6 D-76135 Karlsruhe Tel.: (0721) 86901-17 Fax: (0721) 86901-11 Mobile: (0170) 4464341 Mail: kamenzin(a)netapp.com > -----Original Message----- > From: Premanshu Jain [SMTP:PrJain@shastanets.com] > Sent: Thursday, September 07, 2000 6:15 PM > To: 'Stephen J. Wilcox'; neil lehrer > Cc: toasters > Subject: RE: dlt autoloaders > > a) Why don't you try AIT II instead of DLT. I am getting excellant > performance on those 8 mm. Writing more that 100 GB per tape. We are using > Qualstar and spectra logic tape libraries. > > b) Veritas seems to be OK. With lot of limitations and bugs. I have heard of > a better s/w from www.worksta.com > > c) You may use a Vixel Fibre channal switch with fibre channal interfaces on > the concerened servers and the Filers. This will enable accessing tpaes from > anywhere. > > -----Original Message----- > From: Stephen J. Wilcox [mailto:steve@opaltelecom.co.uk] > Sent: Thursday, September 07, 2000 1:07 AM > To: neil lehrer > Cc: toasters > Subject: Re: dlt autoloaders > > > Actually on that subject I am looking to buy a system soon, I'm thinking > of getting an HP surestore library and using Veritas. Theres a lot of > systems to back up altho they dont have that much data on them, then I > have (for now 1) netfiler, does anyone > > a) think surestore are good? > b) think veritas is good? > c) know if i can do anything to save time by directly connecting the tape > library to the filer hence bypassing the server but still keeping the tape > accessible from the server in order to do my network backups? > > Steve > > -- > Stephen J. Wilcox > Internet Manager, Opal Telecom > http://www.opaltelecom.co.uk/ > Tel: 0161 222 2000 > Fax: 0161 222 2008 > > On Wed, 6 Sep 2000, neil lehrer wrote: > > > hi, > > > > i'd like recommendations for a dlt autoloader to be attached to an f760 > > and controlled by legato. thanks. > > -- > >

1 0

RE: dlt autoloaders
by Premanshu Jain 07 Sep '00

07 Sep '00

a) Why don't you try AIT II instead of DLT. I am getting excellant performance on those 8 mm. Writing more that 100 GB per tape. We are using Qualstar and spectra logic tape libraries. b) Veritas seems to be OK. With lot of limitations and bugs. I have heard of a better s/w from www.worksta.com c) You may use a Vixel Fibre channal switch with fibre channal interfaces on the concerened servers and the Filers. This will enable accessing tpaes from anywhere. -----Original Message----- From: Stephen J. Wilcox [mailto:steve@opaltelecom.co.uk] Sent: Thursday, September 07, 2000 1:07 AM To: neil lehrer Cc: toasters Subject: Re: dlt autoloaders Actually on that subject I am looking to buy a system soon, I'm thinking of getting an HP surestore library and using Veritas. Theres a lot of systems to back up altho they dont have that much data on them, then I have (for now 1) netfiler, does anyone a) think surestore are good? b) think veritas is good? c) know if i can do anything to save time by directly connecting the tape library to the filer hence bypassing the server but still keeping the tape accessible from the server in order to do my network backups? Steve -- Stephen J. Wilcox Internet Manager, Opal Telecom http://www.opaltelecom.co.uk/ Tel: 0161 222 2000 Fax: 0161 222 2008 On Wed, 6 Sep 2000, neil lehrer wrote: > hi, > > i'd like recommendations for a dlt autoloader to be attached to an f760 > and controlled by legato. thanks. > -- >

1 0

Re: gigabit ethernet performance
by Marcus Smiley 07 Sep '00

07 Sep '00

two of 'em (450MHZ), with 2GB RAM. On Wed, 06 Sep 2000, Jim Patterson wrote: ! Marcus, ! ! How many CPUs do you have in your 220R? We have found that a single CPU ! is needed just to keep the network interface busy so the minimum ! configuration to get better than 10MB/sec is two CPUs. ! ! Jim ! -- ! Jim Patterson ! Pfizer Global Research and Development, La Jolla ! ! Marcus Smiley | Ericsson Wireless Communications, Inc. Unix Sys Admin | 6210 Spine Rd. Boulder, CO 80301 | Ph (303) 473-6903

1 0

Re: gigabit ethernet performance
by Darrell Fuhriman 07 Sep '00

07 Sep '00

Premanshu Jain <PrJain(a)shastanets.com> writes: > Oh..That's even better. We got a max of 8 mbytes/sec. Never more than that. > Interestingly, we get the same throughput between any sun boxes as well... No offense, but you must be doing something wrong. Our Suns (E450s) can do at least 60MB/second over gigabit. Darrell

1 0

RE: gigabit ethernet performance
by Premanshu Jain 07 Sep '00

07 Sep '00

Oh..That's even better. We got a max of 8 mbytes/sec. Never more than that. Interestingly, we get the same throughput between any sun boxes as well... prem -----Original Message----- From: Marcus Smiley [mailto:marcus.smiley@ericsson.com] Sent: Wednesday, September 06, 2000 2:41 PM To: toasters(a)mathworks.com Subject: gigabit ethernet performance Does anyone have any benchmarks/suggestions for poor performace with gigabit ethernet between Sun enterprise 220R and NetApp 740 with Intel GBII card? I have an open ticket w/Netapp and it's been open forever. These machines are directly attached (private network), and I've tried a mkfile and copying that file back during the day and at night (no load) and it makes no difference. The performance is ~ 10 MB/s, no matter what. The Sun is running Solaris 8, with latest GB Ethernet 3.0 drivers and patches. Also, the Sun box's only function is Legato server, so I've even stopped all nsr daemons while the test copy/mkfile is in progress, & it makes no difference. I've tried NFSv2 vs. NFSv3, with 32k size packets and 8k size packets. I've tried tcp and udp, with the same results. The /etc/system file on the E220R looks like this, for the ge adapter: * Increase size of STREAMS syncronized queues to increase network * performance. Release notes for qfe card recommends setting this * to 25 per 64Mbytes of RAM in the system. It also prevents * receive overrun on the GbE interface. msmiley 7/20/00 set sq_max_size=775 set ge:ge_adv_1000autoneg_cap=0 set ge:ge_adv_1000fdx_cap=1 set ge:ge_adv_pauseTX=1 Also, I've set tcp/udp specific stuff per the Netapp tech suppt. folks: ndd -set /dev/udp udp_recv_hiwat 65535 ndd -set /dev/udp udp_xmit_hiwat 65535 ndd -set /dev/tcp tcp_recv_hiwat 65535 ndd -set /dev/tcp tcp_xmit_hiwat 65535 The netapp is a 740 (5.3.6R1P1) with over 100GB of free space available on the volume I'm copying to/from. Interestingly enough, I have done a packet trace (pktt) on the GBII card and it looks OK. Also, the throughput from my 100BaseT card to the netapp (thru a catalyst 5500) does ~ 7.5 MB/s. *Any* input here would be much appreciated. -Marc Marcus Smiley | Ericsson Wireless Communications, Inc. Unix Sys Admin | 6210 Spine Rd. Boulder, CO 80301 | Ph (303) 473-6903

1 0

Re: gigabit ethernet performance
by Marcus Smiley 06 Sep '00

06 Sep '00

no there is no switch between the filer and Sun box, using crossover fiber jumper scenario. yes, both machines are set to FULL flow control. Here's a sysstat 1 before, during & after I create a 100MB file, using mkfile: CPU NFS CIFS HTTP Net kB/s Disk kB/s Tape kB/s Cache in out read write read write age 3% 52 0 0 19 43 0 0 0 0 38 3% 65 2 0 11 10 0 0 0 0 38 4% 35 2 0 51 13 0 0 0 0 38 3% 6 23 0 167 10 0 0 0 0 38 4% 36 23 0 14 5 0 0 0 0 38 4% 29 79 0 72 28 52 0 0 0 38 38% 10 260 0 22 57 4412 10316 0 0 38 6% 20 2 0 36 136 3 0 0 0 38 59% 601 1 0 13398 374 0 0 0 0 38 82% 398 16 0 9009 291 4684 14874 0 0 38 73% 572 23 0 12229 363 2972 7208 0 0 38 87% 457 2 0 10112 326 3688 16404 0 0 38 80% 327 0 0 9509 284 5980 14176 0 0 38 73% 520 15 0 14158 405 1093 5053 0 0 38 93% 342 0 0 8070 269 7148 18589 0 0 38 68% 356 2 0 9640 283 5428 11073 0 0 38 82% 591 0 0 11667 346 4112 9208 0 0 38 88% 562 3 0 8520 506 6740 15836 0 0 38 32% 135 1 0 4585 271 1077 4032 0 0 38 3% 19 0 0 4 4 0 0 0 0 38 CPU NFS CIFS HTTP Net kB/s Disk kB/s Tape kB/s Cache in out read write read write age 3% 64 0 0 3 2 0 0 0 0 38 3% 48 0 0 14 12 0 0 0 0 38 3% 17 2 0 13 10 0 0 0 0 38 3% 4 12 0 30 5 0 0 0 0 38 3% 4 12 0 4 2 0 0 0 0 38 On Wed, 06 Sep 2000, Paul Albers wrote: ! Marcus, ! ! Do you have flow control enabled on both Rx and Tx on the ! port the filer is connected to on the switch? If not, try enabling ! it and retest. How does the filer look during writes and ! reads ( sysstat 1 )? Also, what is the configuration of the ! filer ( cards in what slots? ). ! ! Regards, ! ! -paul ! ! Marcus Smiley wrote: ! ! > slot 8 ! > ! > On Wed, 06 Sep 2000, Larry Armstrong wrote: ! > ! > ! Howdy! ! > ! ! > ! Out of curiousity, what slot do you have your GbE card in? ! > ! ! > ! ---larry >:) ! > ! _ ! > ! * Marcus Smiley (marcus.smiley(a)ericsson.com) done spit this rhetoric: ! > ! > Does anyone have any benchmarks/suggestions for poor performace with ! > ! > gigabit ethernet between Sun enterprise 220R and NetApp 740 with Intel ! > ! > GBII card? I have an open ticket w/Netapp and it's been open ! > ! > forever. These machines are directly attached (private network), and ! > ! > I've tried a mkfile and copying that file back during the day and at ! > ! > night (no load) and it makes no difference. ! > ! > ! > ! > The performance is ~ 10 MB/s, no matter what. ! > ! > ! > ! > The Sun is running Solaris 8, with latest GB Ethernet 3.0 drivers ! > ! > and patches. Also, the Sun box's only function is Legato server, ! > ! > so I've even stopped all nsr daemons while the test copy/mkfile ! > ! > is in progress, & it makes no difference. ! > ! > ! > ! > I've tried NFSv2 vs. NFSv3, with 32k size packets and 8k size packets. ! > ! > I've tried tcp and udp, with the same results. ! > ! > ! > ! > The /etc/system file on the E220R looks like this, for the ge adapter: ! > ! > ! > ! > * Increase size of STREAMS syncronized queues to increase network ! > ! > * performance. Release notes for qfe card recommends setting this ! > ! > * to 25 per 64Mbytes of RAM in the system. It also prevents ! > ! > * receive overrun on the GbE interface. msmiley 7/20/00 ! > ! > ! > ! > set sq_max_size=775 ! > ! > set ge:ge_adv_1000autoneg_cap=0 ! > ! > set ge:ge_adv_1000fdx_cap=1 ! > ! > set ge:ge_adv_pauseTX=1 ! > ! > ! > ! > Also, I've set tcp/udp specific stuff per the Netapp tech suppt. folks: ! > ! > ! > ! > ndd -set /dev/udp udp_recv_hiwat 65535 ! > ! > ndd -set /dev/udp udp_xmit_hiwat 65535 ! > ! > ndd -set /dev/tcp tcp_recv_hiwat 65535 ! > ! > ndd -set /dev/tcp tcp_xmit_hiwat 65535 ! > ! > ! > ! > The netapp is a 740 (5.3.6R1P1) with over 100GB of free space available on the ! > ! > volume I'm copying to/from. ! > ! > ! > ! > Interestingly enough, I have done a packet trace (pktt) on the GBII ! > ! > card and it looks OK. Also, the throughput from my 100BaseT card to ! > ! > the netapp (thru a catalyst 5500) does ~ 7.5 MB/s. ! > ! > ! > ! > *Any* input here would be much appreciated. ! > ! > ! > ! > -Marc ! > ! > ! > ! > ! > ! > ! > ! > Marcus Smiley | Ericsson Wireless Communications, Inc. ! > ! > Unix Sys Admin | 6210 Spine Rd. Boulder, CO 80301 | Ph (303) 473-6903 ! > ! > ! > ! ! > ! -- ! > ! Larry Armstrong ! > ! <surfer(a)netapp.com> ! > ! ! > ! bovarism n. - A magnified opinion of one's own abilities. ! > ! > Marcus Smiley | Ericsson Wireless Communications, Inc. ! > Unix Sys Admin | 6210 Spine Rd. Boulder, CO 80301 | Ph (303) 473-6903 ! ! -- ! "Nothing will benefit human health and increase the chances of survival ! of life on earth as much as the evolution to a vegetarian diet." ! ! -Albert Einstein !

1 0

Re: gigabit ethernet performance
by Marcus Smiley 06 Sep '00

06 Sep '00

slot 8 On Wed, 06 Sep 2000, Larry Armstrong wrote: ! Howdy! ! ! Out of curiousity, what slot do you have your GbE card in? ! ! ---larry >:) ! _ ! * Marcus Smiley (marcus.smiley(a)ericsson.com) done spit this rhetoric: ! > Does anyone have any benchmarks/suggestions for poor performace with ! > gigabit ethernet between Sun enterprise 220R and NetApp 740 with Intel ! > GBII card? I have an open ticket w/Netapp and it's been open ! > forever. These machines are directly attached (private network), and ! > I've tried a mkfile and copying that file back during the day and at ! > night (no load) and it makes no difference. ! > ! > The performance is ~ 10 MB/s, no matter what. ! > ! > The Sun is running Solaris 8, with latest GB Ethernet 3.0 drivers ! > and patches. Also, the Sun box's only function is Legato server, ! > so I've even stopped all nsr daemons while the test copy/mkfile ! > is in progress, & it makes no difference. ! > ! > I've tried NFSv2 vs. NFSv3, with 32k size packets and 8k size packets. ! > I've tried tcp and udp, with the same results. ! > ! > The /etc/system file on the E220R looks like this, for the ge adapter: ! > ! > * Increase size of STREAMS syncronized queues to increase network ! > * performance. Release notes for qfe card recommends setting this ! > * to 25 per 64Mbytes of RAM in the system. It also prevents ! > * receive overrun on the GbE interface. msmiley 7/20/00 ! > ! > set sq_max_size=775 ! > set ge:ge_adv_1000autoneg_cap=0 ! > set ge:ge_adv_1000fdx_cap=1 ! > set ge:ge_adv_pauseTX=1 ! > ! > Also, I've set tcp/udp specific stuff per the Netapp tech suppt. folks: ! > ! > ndd -set /dev/udp udp_recv_hiwat 65535 ! > ndd -set /dev/udp udp_xmit_hiwat 65535 ! > ndd -set /dev/tcp tcp_recv_hiwat 65535 ! > ndd -set /dev/tcp tcp_xmit_hiwat 65535 ! > ! > The netapp is a 740 (5.3.6R1P1) with over 100GB of free space available on the ! > volume I'm copying to/from. ! > ! > Interestingly enough, I have done a packet trace (pktt) on the GBII ! > card and it looks OK. Also, the throughput from my 100BaseT card to ! > the netapp (thru a catalyst 5500) does ~ 7.5 MB/s. ! > ! > *Any* input here would be much appreciated. ! > ! > -Marc ! > ! > ! > ! > Marcus Smiley | Ericsson Wireless Communications, Inc. ! > Unix Sys Admin | 6210 Spine Rd. Boulder, CO 80301 | Ph (303) 473-6903 ! > ! ! -- ! Larry Armstrong ! <surfer(a)netapp.com> ! ! bovarism n. - A magnified opinion of one's own abilities. Marcus Smiley | Ericsson Wireless Communications, Inc. Unix Sys Admin | 6210 Spine Rd. Boulder, CO 80301 | Ph (303) 473-6903

1 0

Jump to page:

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