toasters

toasters@lists.teaparty.net

2 participants
13520 discussions

FAS 3210 after upgrade from 8.0.2 to 8.1.3 - higher cache age.
by Rafał Radecki 15 Jul '13

15 Jul '13

Hi All. I've performed an upgrade and afterwards my cache age is higher. Before the upgrade it was ~3 minutes, now it is ~4.5 minutes. The amount of operations over NFS did not change. NFS latency also for all volumes is now 50% of the latency before the upgrade. Cache age with snmp from http://ipmsupport.solarwinds.com/mibs/NETWORK-APPLIANCE-MIB/item.aspx?id=mi… . It is "Age in minutes of the oldest read-only blocks in the buffer cache. This indicates how fast read operations are cycling through system memory; when the appliance is reading very large files (larger than the machine's memory size), buffer cache age will be very low." In the background there is still a rlw_upgrading process running, details in https://communities.netapp.com/thread/22676 . So overall I think that lower nfs latency is better, but am not certain about cache age. Is it not strange both changed so much after the upgrade? Have you had similar experiences? Best regards, Rafal Radecki.

1 0

Netapp ONTAP 8.0.2 - NFS takeover/giveback - should it be visible to clients?
by Rafał Radecki 12 Jul '13

12 Jul '13

Hi All. I have two FAS3210 in 7-mode Actvie/Active. I perform takeover/gieback and see that the outage is visible for clients for nfs mounts. From your experiences should it be so? I think that information about nfs sessions and IP address are moved to the takeover node so in theory it should not be visible. Best regards, Rafal.

10 15

SAS cabling question
by Klise, Steve 09 Jul '13

09 Jul '13

Is there such a thing as SAS patch panels? I may have to go beyond the standard cable lengths to another area of the datacenter.. Has anyone had to get a workaround for this? I think the latest SAS length is 15M or 20M. Thanks Steve

4 3

Snapvault snapshots lost after upgrade.
by Jeff Cleverley 09 Jul '13

09 Jul '13

Greetings, I upgraded heads on our filers. The new 6290s had their own disks and I have 8.1.2P4 on them. The old 6080s were 7.3.5.1P4. For the update I kept the 6290 root/vol0 drives and changed ownership of all the other drives. The aggregates were brought online and all the volumes came up OK. I'm now seeing none of my backups are running because the snapvault snapshot is missing. I still have the acs snapshot. I've tried the snapvault start -S command and it says it is already a snapvault relationship. When I try to tell it to do a snapvault update, it says the base snapshot no longer exists on the source. What happened to all my snapvault snapshots? Why did the snapvault snapshots get released? I thought those were in the volume and should have carried over in the hardware migration. I still have access to the old root volumes and backups of the old root volumes if there is anything that can be salvaged. I would rather not have to re-baseline 100+TB of source data. Thanks, Jeff -- Jeff Cleverley Unix Systems Administrator 4380 Ziegler Road Fort Collins, Colorado 80525 970-288-4611

3 6

Hooking a DS4243 to a "standard" SAS adapter
by Laurent CARON 30 Jun '13

30 Jun '13

Hi, I just decomissionned 2 DS4243. Since I'd rather use them now than let them take the dust, I wonder if it is possible to hook them to a standard SAS adapter (eg: LSI SAS 9207-8e) placed in a server to act as a JBOD enclosure. Physical cabling is ok and will cause no trouble, but will the disks be seen by the card. What about multipath ? Did anybody already do/try it ? Thanks Laurent

2 3

large amount of CIFS other ops from xp desktops after canceled search
by Jordan Slingerland 28 Jun '13

28 Jun '13

Hello toasters, I am having an issue where I periodically find that some xp desktop users are doing a large amount of other ops to a cifs share with a large amount of files for a sustained period. If I let them go, they will continue for days. In most cases, it seems the users do a search and the search takes too long, so they close the window, but it is running in some kind of stuck state in the background. Killing the session or closing open files via connection to the toaster via computer management does nothing; the session just starts back up and pounds the crap out of the filler. Rebooting the desktop has been my solution so far, though I suspect restarting explorer.exe may be sufficient. I am running 8.1.1P2 and this is an IBM rebranded FAS3070. Any ideas? Has anyone seen similar behavior? cifs top -n 10 ops/s reads(n, KB/s) writes(n, KB/s) suspect/s IP Name 1859 | 0 0 | 0 0 | 0 | 10.10.100.223 domain\someuser2 1418 | 0 0 | 0 0 | 0 | 10.10.100.227 domain\someuser1 1673 | 0 0 | 0 0 | 0 | 10.10.180.230 domain\someuser3 qtree status share1 vol1 ntfs enabled normal vol status -v share1 Volume State Status Options Share1 online raid_dp, flex nosnap=off, nosnapdir=off, minra=off, sis no_atime_update=off, nvfail=off, 64-bit ignore_inconsistent=off, snapmirrored=off, create_ucode=on, convert_ucode=off, maxdirsize=73400, schedsnapname=ordinal, fs_size_fixed=on, guarantee=none, svo_enable=off, svo_checksum=off, svo_allow_rman=off, svo_reject_errors=off, no_i2p=off, fractional_reserve=0, extent=off, try_first=volume_grow, read_realloc=off, snapshot_clone_dependency=off, dlog_hole_reserve=off, nbu_archival_snap=off Volume UUID: 0fbdb127-a774-11e2-ab63-123478563412 Containing aggregate: 'sataaggr0' Plex /sataaggr0/plex0: online, normal, active RAID group /sataaggr0/plex0/rg0: normal, block checksums RAID group /sataaggr0/plex0/rg1: normal, block checksums RAID group /sataaggr0/plex0/rg2: normal, block checksums RAID group /sataaggr0/plex0/rg3: normal, block checksums Snapshot autodelete settings for share1: state=off commitment=try trigger=volume target_free_space=20% delete_order=oldest_first defer_delete=user_created prefix=(not specified) destroy_list=none Volume autosize settings: state=off Hybrid Cache: Eligibility=read-write For the fourth consecutive year, Independent Health is the highest ranked health insurance plan in the New York/New Jersey region by J.D. Power and Associates 2013 Member Health Plan StudySM. For the sixth consecutive year, Independent Health was named one of the best companies to work for in New York State. Go to www.independenthealth.com to view careers and available positions. We are an Equal Opportunity Employer. CONFIDENTIALITY NOTICE: This e-mail and its attachments (collectively referred to as "e-mail") may contain confidential information that is privileged and protected from disclosure by Federal and State confidentiality laws, rules or regulations. This e-mail is intended for the designated addressee only. If you are not the designated addressee, you are notified that any disclosure, copying or distribution of this e-mail may be unlawful and may subject you to legal consequences. If you have received this e-mail in error, please contact me immediately by telephone at (716) 631 - 3001 and delete the e-mail from your computer immediately. Thank you for your attention.

2 5

Hardware migration process
by Jeff Cleverley 27 Jun '13

27 Jun '13

Greetings, I think I have a good handle on what I need to do but wanted to check for additional gotchas. We've got some 6290s that will replace some 6080. The 6290s are already built with their own root volume/disks and are running 8.1.2P4. The 6080s are running 7.3.5.1P4. I'm not planning on using the 6080 root disks so I'm not planning on upgrading the 6080s before hand. The plan is basically shut everything down, pull the 7.3.5 root disks, reassign all the other disks to the new system IDs and boot everything up. It is my understanding that there is no 64 bit upgrade to aggregates/volumes that gets done automatically so I'm not seeing and issue there. I have functional /etc/rc, exports, hosts, and username.map files in place on the new hardware. I'm also going to have a copy of the old /etc directories on the new filers so I'll have everything there if I need to go back to them. All needed hardware is in the new heads and supported. The new hardware has newer disk and shelf firmware but I'm going to let it do that during boot up. We have some MOOS drives that have been problematic so I don't want to do those before hand. We have no VM, very little cifs, and 1 SAN/FCP client. Other than that it is all nfs clients. I know I may or may not have to re-join the domain for the cifs piece but I can deal with that. It seems this should be relatively simple and easy, but that's usually when I get into trouble :-) Any obvious holes/flaws in my plans that I should be aware of? Thanks, Jeff -- Jeff Cleverley Unix Systems Administrator 4380 Ziegler Road Fort Collins, Colorado 80525 970-288-4611

3 11

FWD: Any reason not to buy a FAS3220 over a FAS2240-2?
by Matt Kilham 27 Jun '13

27 Jun '13

Thanks to all those who replied - very helpful. We're going to go for te 3220 with FlashPool. Cheers, Matt Original Message: ---------- Forwarded Message ---------- FROM: Matt Kilham (stratton) TO: Mailing Lists DATE: Thu, 27 Jun 2013 12:32:42 +1000 RE: Any reason not to buy a FAS3220 over a FAS2240-2? Hi everyone, We are getting towards the pointy end of the procurement phase of an infrastructure refresh at our SME. Part of this is replacing an aging (badly) FAS2050 + a couple of DS14Mk4 shelves with a new NetApp filer. The FAS2240-2 was an obvious candidate, as even an entry-level system such as this provides more than enough performance and capacity for our current and planned needs for the next few years. However, a key part of our upgrade plan was implementing FlashPool, and we only discovered late in the piece that the FAS2xxx series has a 400GB usable FlashPool limit per HA pair. This is/was cause for concern around whether this would be sufficient for us given a pending VDI deployment, and led to a flurry of research into alternatives, quotes on alternative systems, etc. Where we have "landed" is that we have been quoted on two options - the original FAS2240-2 and a FAS3220*, with otherwise identical disks/software licenses/etc. The interesting part is that the FAS3220 quote is only fractionally more expensive than the 2240-2 (think 5% or so). I am being told that the reason for the close-to-price-parity is a sharp deal on the FAS3220, whereas the FAS2240-2 is close to being a fixed price / fixed margin unit. While I have not yet verified the truth of the above (that the reason for price parity is a cheap 3220 quote, not an expensive 2240-2 quote), if we take it at face value that this is true and hence can buy a FAS3220 for essentially the same price as a FAS2240-2, can anyone think of any good reasons why we wouldn't take the FAS3220 (remembering it is overkill for us in every way except the FlashPool limit)? All I can think of is that ongoing support / warranty will be a bit more expensive, and any additional software licenses purchased later will cost more (can anyone quantify approx how much?). I think the former is worth the cost, and the latter isn't too much of a concern as we barely use any separately licensed features now and I don't envision this changing in the future (except perhaps for SnapMirror - we may implement that). Thoughts from the gurus? And thanks, in advance. * we actually have two FAS3220 quotes, which are essentially the same price and spec except that one provides 7 x 200GB SSDs for 400GB + 200GB of FlashPool, and the other provides 2 x 512GB FlashCache cards. Secondary question here: for a small environment like ours that will only have one aggregate per controller, I'm thinking the FlashPool options makes more sense, but would like sanity check here - thoughts? Cheers, Matt

1 0

Any reason not to buy a FAS3220 over a FAS2240-2?
by Matt Kilham 27 Jun '13

27 Jun '13

Hi everyone, We are getting towards the pointy end of the procurement phase of an infrastructure refresh at our SME. Part of this is replacing an aging (badly) FAS2050 + a couple of DS14Mk4 shelves with a new NetApp filer. The FAS2240-2 was an obvious candidate, as even an entry-level system such as this provides more than enough performance and capacity for our current and planned needs for the next few years. However, a key part of our upgrade plan was implementing FlashPool, and we only discovered late in the piece that the FAS2xxx series has a 400GB usable FlashPool limit per HA pair. This is/was cause for concern around whether this would be sufficient for us given a pending VDI deployment, and led to a flurry of research into alternatives, quotes on alternative systems, etc. Where we have "landed" is that we have been quoted on two options - the original FAS2240-2 and a FAS3220*, with otherwise identical disks/software licenses/etc. The interesting part is that the FAS3220 quote is only fractionally more expensive than the 2240-2 (think 5% or so). I am being told that the reason for the close-to-price-parity is a sharp deal on the FAS3220, whereas the FAS2240-2 is close to being a fixed price / fixed margin unit. While I have not yet verified the truth of the above (that the reason for price parity is a cheap 3220 quote, not an expensive 2240-2 quote), if we take it at face value that this is true and hence can buy a FAS3220 for essentially the same price as a FAS2240-2, can anyone think of any good reasons why we wouldn't take the FAS3220 (remembering it is overkill for us in every way except the FlashPool limit)? All I can think of is that ongoing support / warranty will be a bit more expensive, and any additional software licenses purchased later will cost more (can anyone quantify approx how much?). I think the former is worth the cost, and the latter isn't too much of a concern as we barely use any separately licensed features now and I don't envision this changing in the future (except perhaps for SnapMirror - we may implement that). Thoughts from the gurus? And thanks, in advance. * we actually have two FAS3220 quotes, which are essentially the same price and spec except that one provides 7 x 200GB SSDs for 400GB + 200GB of FlashPool, and the other provides 2 x 512GB FlashCache cards. Secondary question here: for a small environment like ours that will only have one aggregate per controller, I'm thinking the FlashPool options makes more sense, but would like sanity check here - thoughts? Cheers, Matt

1 0

6290 10G ports vs PCI 10G cards
by Jeff Cleverley 19 Jun '13

19 Jun '13

Greetings, I haven't been able to find out anything about the bus speeds for the ports on a 6290. The hardware universe shows an order to fill slots for things like 10G or SAS controllers. This makes me believe all the ports are not the same. I was thinking of using the on-board 10G ports instead of a dual port PCI card that would be in one of the top 5 priority slots. We do have the IOX module on the arrays. Does anyone know where I can find this information? Thanks, Jeff -- Jeff Cleverley Unix Systems Administrator 4380 Ziegler Road Fort Collins, Colorado 80525 970-288-4611

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