Now that the R100 has come out I was wondering what people think of the seperate vol0 concept? In the old days it wasn't quite so bad; ok, so you wasted a 4gb drive or maybe even an 18gb drive. But now we're looking at tossing a 136gb drive for what? Squat really.
So what's the new thought on this? Do you not waste the disk on a seperate vol0 and just keep tight reigns on qtrees? Or is there still an argument for vol0 on it's own?
Thanks.
Jeff, I personally don't like active data on vol0. In the past I have put archive data on vol0 because it is low impact. I happen to be a little more paranoid than some but that is just me and we have been bitten before by putting highly volatile on the root volume. Chris
On Mon, Feb 04, 2002 at 02:02:14PM -0800, Jeff Kennedy wrote:
Now that the R100 has come out I was wondering what people think of the seperate vol0 concept? In the old days it wasn't quite so bad; ok, so you wasted a 4gb drive or maybe even an 18gb drive. But now we're looking at tossing a 136gb drive for what? Squat really.
So what's the new thought on this? Do you not waste the disk on a seperate vol0 and just keep tight reigns on qtrees? Or is there still an argument for vol0 on it's own?
Thanks.
===================== Jeff Kennedy Unix Administrator AMCC jlkennedy@amcc.com
I don't disagree but this seems to be a paradigm of the past. With the R100 out, and OT 6.2 comming soon, the limitations are quickly disappearing. Currently I plan to just snapmirror volumes to the R100, but with 6.2 I plan to create volumes in the 1TB+ range and snapmirror the qtrees (everything we do is in qtrees). I used to be a volume nazi for probably the same reasons you mention (no volume was to *ever* get above 400GB's non-snapshot), now I don't see that as a problem anymore. I have no issues with volumes going past 1TB if this works as advertised, and I have no reason to think it won't.
So, do you still think vol0 should be seperate? And don't give me the "That's the way we've always done it..." routine :-] Keep in mind the R100 does mostly archival, backup/restore, and misc funtions like db exports and IT tool storage, not traditional production stuff.
~JK
Chris Blackmor wrote:
Jeff, I personally don't like active data on vol0. In the past I have put archive data on vol0 because it is low impact. I happen to be a little more paranoid than some but that is just me and we have been bitten before by putting highly volatile on the root volume. Chris
On Mon, Feb 04, 2002 at 02:02:14PM -0800, Jeff Kennedy wrote:
Now that the R100 has come out I was wondering what people think of the seperate vol0 concept? In the old days it wasn't quite so bad; ok, so you wasted a 4gb drive or maybe even an 18gb drive. But now we're looking at tossing a 136gb drive for what? Squat really.
So what's the new thought on this? Do you not waste the disk on a seperate vol0 and just keep tight reigns on qtrees? Or is there still an argument for vol0 on it's own?
Thanks.
===================== Jeff Kennedy Unix Administrator AMCC jlkennedy@amcc.com
--
| When you're throwin' your *- Chris Blackmor _______ | weight around, *
- Advanced Micro Devices ____ | | be ready to have it *
- Phone: (512) 602-1608 /| | | | thrown around by *
- Fax: (512) 602-5155 | |___| | | somebody else. *
- Email: chris.blackmor@amd.com |____/ | | Author Unknown*
My comments are mine, and mine alone. *
We don't have an R100 on site. As far as size of volume goes, until I let someone else bleed on snapshots by qtrees (backups by qtrees) we probably won't go over 750G volumes. With restores of volumes taking in the area of 30G/hr in our configuration we don't want to risk outages of more than 24 hours if we lose a volume. This is a risk based decision.
As far as data on vol0... we actually used to ALWAYS use vol0 for active data. After being bit some we went to low impact data over the last 2 years. Do I disagree or agree... I will take the easy way out and say that this would really be site dependent. The type of data one site has vs the type of data another site has could drastically change the way in which we might manage our data. I understand what you are saying about qtrees and 6.2. I think that eventually I will let lose of the reins and allow it again. Maybe. ;)
C-
On Tue, Feb 05, 2002 at 06:12:32AM -0800, Jeff Kennedy wrote:
I don't disagree but this seems to be a paradigm of the past. With the R100 out, and OT 6.2 comming soon, the limitations are quickly disappearing. Currently I plan to just snapmirror volumes to the R100, but with 6.2 I plan to create volumes in the 1TB+ range and snapmirror the qtrees (everything we do is in qtrees). I used to be a volume nazi for probably the same reasons you mention (no volume was to *ever* get above 400GB's non-snapshot), now I don't see that as a problem anymore. I have no issues with volumes going past 1TB if this works as advertised, and I have no reason to think it won't.
So, do you still think vol0 should be seperate? And don't give me the "That's the way we've always done it..." routine :-] Keep in mind the R100 does mostly archival, backup/restore, and misc funtions like db exports and IT tool storage, not traditional production stuff.
~JK
Chris Blackmor wrote:
Jeff, I personally don't like active data on vol0. In the past I have put archive data on vol0 because it is low impact. I happen to be a little more paranoid than some but that is just me and we have been bitten before by putting highly volatile on the root volume. Chris
On Mon, Feb 04, 2002 at 02:02:14PM -0800, Jeff Kennedy wrote:
Now that the R100 has come out I was wondering what people think of the seperate vol0 concept? In the old days it wasn't quite so bad; ok, so you wasted a 4gb drive or maybe even an 18gb drive. But now we're looking at tossing a 136gb drive for what? Squat really.
So what's the new thought on this? Do you not waste the disk on a seperate vol0 and just keep tight reigns on qtrees? Or is there still an argument for vol0 on it's own?
Thanks.
===================== Jeff Kennedy Unix Administrator AMCC jlkennedy@amcc.com
--
| When you're throwin' your *- Chris Blackmor _______ | weight around, *
- Advanced Micro Devices ____ | | be ready to have it *
- Phone: (512) 602-1608 /| | | | thrown around by *
- Fax: (512) 602-5155 | |___| | | somebody else. *
- Email: chris.blackmor@amd.com |____/ | | Author Unknown*
My comments are mine, and mine alone. *
--
Jeff Kennedy Unix Administrator AMCC jlkennedy@amcc.com
An argument for isolating root volume is that we are still warned in the netapp admin classes that if the root volume gets full, the filer will crash. We can build qtress to limit client to just under 100% or we can reserve 1% snap and turn off snaps. But will either of those really work? Will the filer potentially crash on a full volume even if it is not root volume?
At 6:12 AM -0800 2/5/02, Jeff Kennedy wrote:
I don't disagree but this seems to be a paradigm of the past. With the R100 out, and OT 6.2 comming soon, the limitations are quickly disappearing. Currently I plan to just snapmirror volumes to the R100, but with 6.2 I plan to create volumes in the 1TB+ range and snapmirror the qtrees (everything we do is in qtrees). I used to be a volume nazi for probably the same reasons you mention (no volume was to *ever* get above 400GB's non-snapshot), now I don't see that as a problem anymore. I have no issues with volumes going past 1TB if this works as advertised, and I have no reason to think it won't.
So, do you still think vol0 should be seperate? And don't give me the "That's the way we've always done it..." routine :-] Keep in mind the R100 does mostly archival, backup/restore, and misc funtions like db exports and IT tool storage, not traditional production stuff.
~JK
Chris Blackmor wrote:
Jeff, I personally don't like active data on vol0. In the past I have put archive data on vol0 because it is low impact. I happen to be a little more paranoid than some but that is just me and we have been bitten before by putting highly volatile on the root volume. Chris
On Mon, Feb 04, 2002 at 02:02:14PM -0800, Jeff Kennedy wrote:
Now that the R100 has come out I was wondering what people think of the seperate vol0 concept? In the old days it wasn't quite so bad; ok, so you wasted a 4gb drive or maybe even an 18gb drive. But now we're looking at tossing a 136gb drive for what? Squat really.
So what's the new thought on this? Do you not waste the disk on a seperate vol0 and just keep tight reigns on qtrees? Or is there still an argument for vol0 on it's own?
Thanks.
===================== Jeff Kennedy Unix Administrator AMCC jlkennedy@amcc.com
--
| When you're throwin' your *- Chris Blackmor _______ | weight around, *
- Advanced Micro Devices ____ | | be ready to have it *
- Phone: (512) 602-1608 /| | | | thrown around by *
- Fax: (512) 602-5155 | |___| | | somebody else. *
- Email: chris.blackmor@amd.com |____/ | | Author Unknown*
My comments are mine, and mine alone. *
--
Jeff Kennedy Unix Administrator AMCC jlkennedy@amcc.com
I don't know about the 1% snapshot reserve idea, but keeping the qtree quotas to under 100%-vol0 does work. If vol0 fills up the filer will crash, if any other volume fills up the filer will continue, so this is a definite argument for seperate vol0. However, the R100 is primarily archive, backup/restore, snapvault, et al, so if it bombed for a bit for any reason it should not have a severe impact. This is why the shelves only have 1 PS (it's true, look for yourself). There is not the same redundancy built into this box for just this reason, it's a non-production, archival type of system. The kind that could afford to go down every so often just for fun...
~JK
Jim Harm wrote:
An argument for isolating root volume is that we are still warned in the netapp admin classes that if the root volume gets full, the filer will crash. We can build qtress to limit client to just under 100% or we can reserve 1% snap and turn off snaps. But will either of those really work? Will the filer potentially crash on a full volume even if it is not root volume?
At 6:12 AM -0800 2/5/02, Jeff Kennedy wrote:
I don't disagree but this seems to be a paradigm of the past. With the R100 out, and OT 6.2 comming soon, the limitations are quickly disappearing. Currently I plan to just snapmirror volumes to the R100, but with 6.2 I plan to create volumes in the 1TB+ range and snapmirror the qtrees (everything we do is in qtrees). I used to be a volume nazi for probably the same reasons you mention (no volume was to *ever* get above 400GB's non-snapshot), now I don't see that as a problem anymore. I have no issues with volumes going past 1TB if this works as advertised, and I have no reason to think it won't.
So, do you still think vol0 should be seperate? And don't give me the "That's the way we've always done it..." routine :-] Keep in mind the R100 does mostly archival, backup/restore, and misc funtions like db exports and IT tool storage, not traditional production stuff.
~JK
Chris Blackmor wrote:
Jeff, I personally don't like active data on vol0. In the past I have put archive data on vol0 because it is low impact. I happen to be a little more paranoid than some but that is just me and we have been bitten before by putting highly volatile on the root volume. Chris
On Mon, Feb 04, 2002 at 02:02:14PM -0800, Jeff Kennedy wrote:
Now that the R100 has come out I was wondering what people think of the seperate vol0 concept? In the old days it wasn't quite so bad; ok, so you wasted a 4gb drive or maybe even an 18gb drive. But now we're looking at tossing a 136gb drive for what? Squat really.
So what's the new thought on this? Do you not waste the disk on a seperate vol0 and just keep tight reigns on qtrees? Or is there still an argument for vol0 on it's own?
Thanks.
===================== Jeff Kennedy Unix Administrator AMCC jlkennedy@amcc.com
--
| When you're throwin' your *- Chris Blackmor _______ | weight around, *
- Advanced Micro Devices ____ | | be ready to have it *
- Phone: (512) 602-1608 /| | | | thrown around by *
- Fax: (512) 602-5155 | |___| | | somebody else. *
- Email: chris.blackmor@amd.com |____/ | | Author Unknown*
My comments are mine, and mine alone. *
--
Jeff Kennedy Unix Administrator AMCC jlkennedy@amcc.com
-- }}}===============>> LLNL James E. Harm (Jim); jharm@llnl.gov System Administrator Compaq Clusters (925) 422-4018 Page: 423-7705x57152
Qtrees limiting space to under 100% will work. One must ensure that qtree are always on for vol0 though. Reserving snapshot space does not resolve this problem as you may still fill up the volume to 100%. It will create a quick out for you to remove the snap reserve but will not prevent the volume from filling up. Myself, I prefer to isolate the root volume to only system activities with no user data.
-gdg
Jim Harm wrote:
An argument for isolating root volume is that we are still warned in the netapp admin classes that if the root volume gets full, the filer will crash. We can build qtress to limit client to just under 100% or we can reserve 1% snap and turn off snaps. But will either of those really work? Will the filer potentially crash on a full volume even if it is not root volume?
At 6:12 AM -0800 2/5/02, Jeff Kennedy wrote:
I don't disagree but this seems to be a paradigm of the past. With the R100 out, and OT 6.2 comming soon, the limitations are quickly disappearing. Currently I plan to just snapmirror volumes to the R100, but with 6.2 I plan to create volumes in the 1TB+ range and snapmirror the qtrees (everything we do is in qtrees). I used to be a volume nazi for probably the same reasons you mention (no volume was to *ever* get above 400GB's non-snapshot), now I don't see that as a problem anymore. I have no issues with volumes going past 1TB if this works as advertised, and I have no reason to think it won't.
So, do you still think vol0 should be seperate? And don't give me the "That's the way we've always done it..." routine :-] Keep in mind the R100 does mostly archival, backup/restore, and misc funtions like db exports and IT tool storage, not traditional production stuff.
~JK
Chris Blackmor wrote:
Jeff, I personally don't like active data on vol0. In the past I have put archive data on vol0 because it is low impact. I happen to be a little more paranoid than some but that is just me and we have been bitten before by putting highly volatile on the root volume. Chris
On Mon, Feb 04, 2002 at 02:02:14PM -0800, Jeff Kennedy wrote:
Now that the R100 has come out I was wondering what people think of the seperate vol0 concept? In the old days it wasn't quite so bad; ok, so you wasted a 4gb drive or maybe even an 18gb drive. But now we're looking at tossing a 136gb drive for what? Squat really.
So what's the new thought on this? Do you not waste the disk on a seperate vol0 and just keep tight reigns on qtrees? Or is there still an argument for vol0 on it's own?
Thanks.
===================== Jeff Kennedy Unix Administrator AMCC jlkennedy@amcc.com
--
| When you're throwin' your *- Chris Blackmor _______ | weight around, *
- Advanced Micro Devices ____ | | be ready to have it *
- Phone: (512) 602-1608 /| | | | thrown around by *
- Fax: (512) 602-5155 | |___| | | somebody else. *
- Email: chris.blackmor@amd.com |____/ | | Author Unknown*
My comments are mine, and mine alone. *
--
Jeff Kennedy Unix Administrator AMCC jlkennedy@amcc.com
-- }}}===============>> LLNL James E. Harm (Jim); jharm@llnl.gov System Administrator Compaq Clusters (925) 422-4018 Page: 423-7705x57152
-
Chris Blackmor -
Jeff Kennedy -
Jim Harm -
Mr. G. D. Geen