toasters

toasters@lists.teaparty.net

13514 discussions

RE: Double Parity on an Oracle Volume
by Jack Lyons 29 Jul '04

29 Jul '04

Ditto....although we deal with just CIFS file access, we did some pretty intense testing when the system was installed and after any change to the configuration. After we moved to DP we didn't notice a significant change in performance. Jack -----Original Message----- From: Skottie Miller [mailto:skottie@anim.dreamworks.com] Sent: Wednesday, July 28, 2004 2:51 PM To: Robert Borowicz Cc: toasters(a)mathworks.com Subject: Re: Double Parity on an Oracle Volume Robert Borowicz wrote: > I'm planning out the next two years of our Database Storage on my FAS940 > and the DBA found out about the NetApps Double Parity feature and thinks > we should implement it. > > My first concern is performance. Is there a performance hit when you > implement DP? I switched all our Oracle DB volumes to raid_dp, and for our workload, did not see any measureable change in performance. The extra warm fuzzy that goes with the additional protection makes people happy. Our production Oracle data spaces are on fas980's -skottie -- Scott Miller | Dreamworks Animation work: skottie(a)dreamworks.com | life: skottie(a)pobox.com | http://skottie-di.net This email and its contents may be confidential. If it is and you are not the intended recipient, please do not disclose or use the information within this email or its attachments. If you have received this email in error, please delete it immediately. Thank you.

1 0

Hey There
by pam66aK＠yahoo.com 29 Jul '04

29 Jul '04

it's me pamla ;) zzzzzz i'm so tired today, been working alot lately... I just realized since I've been away for sometime, something happend to my website, maybe that's why I haven't heard from you. Sorry, but I don't know about computers much :/ Just got home, took a hot bath ;) and now feel like chatting with someone. Hopefully everything works this time, i put everything on free online space at http://camhost4free.info/pamz/ , try to email me when you get a chance ... h0t kiss, pamela -)

1 0

CIFS client slowness
by Brubaker, Paul 29 Jul '04

29 Jul '04

We're experiencing severe latency when a Windows 2000pro or Windows XP client reads and writes to a folder on our Netapps (FAS960 w/ Data Ontap 6.5.1). It only seems to be client O/S as our testing w/ a Windows 2000 Server did not have latency on reads / writes. We're currently looking at SACK and TCP Window sizes. Has anyone experienced severe client latency? Any know issues / advice? Anyone changing the sack option on their Filers? Thanks! Related info below: ===================== SackOpts This parameter controls whether or not SACK (Selective Acknowledgement) support is enabled, as specified in RFC 2018. SACK is especially important for connections using large TCP Window sizes. HKLM\SYSTEM\CurrentControlSet\Services\Tcpip\Parameters SackOpts="1" (DWORD - boolean, recommended setting is 1. Possible settings are 0 - No Sack options or 1 - Sack Option enabled). http://www.speedguide.net/read_articles.php?id=157 <http://www.speedguide.net/read_articles.php?id=157> =================== One more item of note. In the traces from the slow xp box, we're showing a boatload of SACK transmissions, whereas in the fast one, not a one. I'd like to turn off SACK on the filer side to see if things clear up or not as google isn't giving me much luck in the 'which registry entry do I change to check things on the windows box' angle. On the filer: filer> options ip.tcp.sack.enable off Paul M. Brubaker, Jr. RIDS - Intel System Support AT&T Wireless - Harrisburg PA e-mail: paul.brubaker(a)attws.com office #: 717-526-5011 cell #: 717-578-2254

2 1

Someone is looking for experiences about: toasters@mathworks.com
by SYEC Support Department 29 Jul '04

29 Jul '04

********This is not commercial email.******** ALERT! - A member is trying to share Opinions and Experiences about you in our online community. The purpose of this email is to inform you that a submission has been made about you at our website. This is email is not commercial in nature. If this email message was delivered to your spam or bulk email folder please notify your ISP or spam filtering company regarding this mistake on their part. To view submissions about you click here: http://4.share4.biz/lx.php?a=search&b=5&c=toasters@mathworks.com This website also includes a highly valuable Daily Searching System. This is a simple system in which you can set-up searches that this website will perform for you each and every day. After performing the searches that you have specified, our Daily Searching System will send your search results to you via email. IMPORTANT - To avoid future notifications like this simply add this email address to our Do Not Email List here: http://7.share2.info/lx.php?a=donotemail&b=toasters@mathworks.com Sincerely, SYEC Support

1 0

Double Parity on an Oracle Volume
by Robert Borowicz 28 Jul '04

28 Jul '04

I'm planning out the next two years of our Database Storage on my FAS940 and the DBA found out about the NetApps Double Parity feature and thinks we should implement it. My first concern is performance. Is there a performance hit when you implement DP? Thanks in advance. -Rob

2 1

RE: CIFS client slowness
by Brubaker, Paul 28 Jul '04

28 Jul '04

Correction... we have both Ontap 6.4.2 and 6.5.1p1 running... both FAS960s have CIFS latency issues. And we have an open case w/ Netapp... Paul M. Brubaker, Jr. RIDS - Intel System Support AT&T Wireless - Harrisburg PA e-mail: paul.brubaker(a)attws.com office #: 717-526-5011 cell #: 717-578-2254 -----Original Message----- From: owner-toasters(a)mathworks.com [mailto:owner-toasters@mathworks.com] On Behalf Of Brubaker, Paul Sent: Wednesday, July 28, 2004 12:43 PM To: toasters(a)mathworks.com Subject: CIFS client slowness We're experiencing severe latency when a Windows 2000pro or Windows XP client reads and writes to a folder on our Netapps (FAS960 w/ Data Ontap 6.5.1). It only seems to be client O/S as our testing w/ a Windows 2000 Server did not have latency on reads / writes. We're currently looking at SACK and TCP Window sizes. Has anyone experienced severe client latency? Any know issues / advice? Anyone changing the sack option on their Filers? Thanks! Related info below: ===================== SackOpts This parameter controls whether or not SACK (Selective Acknowledgement) support is enabled, as specified in RFC 2018. SACK is especially important for connections using large TCP Window sizes. HKLM\SYSTEM\CurrentControlSet\Services\Tcpip\Parameters SackOpts="1" (DWORD - boolean, recommended setting is 1. Possible settings are 0 - No Sack options or 1 - Sack Option enabled). http://www.speedguide.net/read_articles.php?id=157 <http://www.speedguide.net/read_articles.php?id=157> =================== One more item of note. In the traces from the slow xp box, we're showing a boatload of SACK transmissions, whereas in the fast one, not a one. I'd like to turn off SACK on the filer side to see if things clear up or not as google isn't giving me much luck in the 'which registry entry do I change to check things on the windows box' angle. On the filer: filer> options ip.tcp.sack.enable off Paul M. Brubaker, Jr. RIDS - Intel System Support AT&T Wireless - Harrisburg PA e-mail: paul.brubaker(a)attws.com office #: 717-526-5011 cell #: 717-578-2254

1 0

unable to log in
by Stephenson, Burke T. GD 28 Jul '04

28 Jul '04

I have a new FAS960 and I am migrating all users and data from F740 to FAS960. I have the data moved and ready to change over. To test it I changed a test user account in the NIS maps to point to the new filer and location. When I try to log in using a UNIX box it just kicks me right back out almost as if it can't write to the home directory. - I have checked all of the options, usermap.cfg, passwd, quotas, exports, files and they are all identical and I can not see any discrepancies in the /etc directory between the two filers. - The options are all the same also. - In my NIS maps I can ypcat the auto.home and the test user account lists the correct filer and home directory. I just can not log in. Netgroup is correct. - I have cifs.trace_login turned on and I don't even see it attempting to log in. - My cifs shares are correct. - The filer is correctly in the domain (win2k) and also resolves using NIS. I can run a cifs testdc successfully and cifs domaininfo shows correct info - I can mount a share and home directory - I can map a share from win2k - when I telnet into another machine as the test user it maps to the new home directory correctly but when I try to touch a file it says it can not create the file, even as root. - qtree security is set to unix It is almost as if the volume is read only but it is not. I just do not know where to look now. I hope that I am just overlooking something simple. Burke

2 1

RE: Zeroing out disks
by Strange, Steve 27 Jul '04

27 Jul '04

Is there any reason NOT to zero disks that you are leaving in the system as spares? They will have to be zeroed when you add them to a volume (or create a new one) anyway, thus delaying their availability in that volume for data. If they are chosen for reconstruct, they do not need to be zeroed ahead of time. But either way, the performance effect on the system of zeroing disks is negligible, so why not zero them all? Cheers, Steve -----Original Message----- From: George Kahler [mailto:george@YorkU.CA] Sent: Tuesday, July 27, 2004 10:15 AM To: watherton(a)BERKCOM.com Cc: toasters(a)mathworks.com Subject: Re: Zeroing out disks Thanks, that is exactly what I wanted. It is too bad that one has to go round about way of accomplishing this. An argument to the spares_zero command would do the trick. I have few disks (now spares) that used to belong to a volume that was destroyed, however, I MUST zero out any data from these disks before they can leave our premises. George ------------------------------------------------------------------------ ------- George Kahler e-mail: george(a)yorku.ca Sr. Systems Administrator humans: (416) 736-2100 x.22699 Computing and Network Services machines: (416) 736-5830 Ontario, Canada, M3J-1P3 On Tue, 27 Jul 2004 10:01:27 -0700, William Atherton <watherton(a)BERKCOM.com> wrote: > George, > > The NetApp Zeros the disks in parallel so the time it takes to zero 1 disk > should be the same as 7 disks. If you really do not want to zero the other > spares out you could pull them from the filer and then run the command to > zero all and then place the none zeroed disk back in. Hope this helps. > > Thank you, > > William Atherton > > watherton(a)berkcom.com > 510-644-1599 VOICE > 510-644-1598 FAX > 888-812-9040 TOLL-FREE > > Berkeley Communications Corporation > 2990 San Pablo Avenue > Berkeley, CA 94702 > http://www.berkcom.com > > > -----Original Message----- > From: George Kahler [mailto:george@YorkU.CA] > Sent: Tuesday, July 27, 2004 6:53 AM > To: toasters(a)mathworks.com > Subject: Zeroing out disks > > > Does anyone know how you can zero out a specific disk(s) ? > The "disk zero spares" command or "spares_zero" diagnostic command > seem to zero out ALL spare disks only. > > Thanks, > George

5 5

RE: Zeroing out disks
by Strange, Steve 27 Jul '04

27 Jul '04

The "disk zero spares" command will only zero spares that aren't already zeroed. Steve -----Original Message----- From: George Kahler [mailto:george@yorku.ca] Sent: Tuesday, July 27, 2004 10:47 AM To: Strange, Steve Cc: watherton(a)BERKCOM.com; toasters(a)mathworks.com Subject: Re: Zeroing out disks Becase most of the other spares if not all are already zeroed. George On Tue, 27 Jul 2004 10:40:15 -0700, "Strange, Steve" <Steve.Strange(a)netapp.com> wrote: > Is there any reason NOT to zero disks that you are leaving in the system > as spares? They will have to be zeroed when you add them to a volume > (or create a new one) anyway, thus delaying their availability in that > volume for data. If they are chosen for reconstruct, they do not need > to be zeroed ahead of time. But either way, the performance effect on > the system of zeroing disks is negligible, so why not zero them all? > > Cheers, > > Steve > > -----Original Message----- > From: George Kahler [mailto:george@YorkU.CA] > Sent: Tuesday, July 27, 2004 10:15 AM > To: watherton(a)BERKCOM.com > Cc: toasters(a)mathworks.com > Subject: Re: Zeroing out disks > > Thanks, > that is exactly what I wanted. It is too bad that one has to go round > about way of accomplishing this. An argument to the spares_zero > command would do the trick. > > I have few disks (now spares) that used to belong to a volume that was > destroyed, however, I MUST zero out any data from these disks before > they can leave our premises. > > George > > ------------------------------------------------------------------------ > ------- > George Kahler e-mail: george(a)yorku.ca > > Sr. Systems Administrator humans: (416) 736-2100 > x.22699 > Computing and Network Services machines: (416) 736-5830 > Ontario, Canada, M3J-1P3 > > On Tue, 27 Jul 2004 10:01:27 -0700, William Atherton > <watherton(a)BERKCOM.com> wrote: > > George, > > > > The NetApp Zeros the disks in parallel so the time it takes to zero 1 > disk > > should be the same as 7 disks. If you really do not want to zero the > other > > spares out you could pull them from the filer and then run the command > to > > zero all and then place the none zeroed disk back in. Hope this helps. > > > > Thank you, > > > > William Atherton > > > > watherton(a)berkcom.com > > 510-644-1599 VOICE > > 510-644-1598 FAX > > 888-812-9040 TOLL-FREE > > > > Berkeley Communications Corporation > > 2990 San Pablo Avenue > > Berkeley, CA 94702 > > http://www.berkcom.com > > > > > > -----Original Message----- > > From: George Kahler [mailto:george@YorkU.CA] > > Sent: Tuesday, July 27, 2004 6:53 AM > > To: toasters(a)mathworks.com > > Subject: Zeroing out disks > > > > > > Does anyone know how you can zero out a specific disk(s) ? > > The "disk zero spares" command or "spares_zero" diagnostic command > > seem to zero out ALL spare disks only. > > > > Thanks, > > George

1 0

Re: Zeroing out disks
by George Kahler 27 Jul '04

27 Jul '04

Thanks, that is exactly what I wanted. It is too bad that one has to go round about way of accomplishing this. An argument to the spares_zero command would do the trick. I have few disks (now spares) that used to belong to a volume that was destroyed, however, I MUST zero out any data from these disks before they can leave our premises. George ------------------------------------------------------------------------------- George Kahler e-mail: george(a)yorku.ca Sr. Systems Administrator humans: (416) 736-2100 x.22699 Computing and Network Services machines: (416) 736-5830 Ontario, Canada, M3J-1P3 On Tue, 27 Jul 2004 10:01:27 -0700, William Atherton <watherton(a)BERKCOM.com> wrote: > George, > > The NetApp Zeros the disks in parallel so the time it takes to zero 1 disk > should be the same as 7 disks. If you really do not want to zero the other > spares out you could pull them from the filer and then run the command to > zero all and then place the none zeroed disk back in. Hope this helps. > > Thank you, > > William Atherton > > watherton(a)berkcom.com > 510-644-1599 VOICE > 510-644-1598 FAX > 888-812-9040 TOLL-FREE > > Berkeley Communications Corporation > 2990 San Pablo Avenue > Berkeley, CA 94702 > http://www.berkcom.com > > > -----Original Message----- > From: George Kahler [mailto:george@YorkU.CA] > Sent: Tuesday, July 27, 2004 6:53 AM > To: toasters(a)mathworks.com > Subject: Zeroing out disks > > > Does anyone know how you can zero out a specific disk(s) ? > The "disk zero spares" command or "spares_zero" diagnostic command > seem to zero out ALL spare disks only. > > Thanks, > George

2 1

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