Options
- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Mute
- Printer Friendly Page
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
did any see thw case where volume ran out of inodes.
-- did you increase the vol size or
just increased the inode count, if so any perf issues.
Solved! See The Solution
1 ACCEPTED SOLUTION
migration has accepted the solution
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hi Reddy,
Simple concept: Each 4K block is 1 node. lets say you have 1TB volume, the number of inodes you can have is: (1024 * 1024 *1024)/4= 268435456 inodes. The best practice is use upto 75 to 85% of nodes of a volume. If you further need the inodes, you better increase the size of the volume and increase the inodes after the threshold is reached (85% inodes consumption).
First 1024 is GB and second 1024 is mb and third 1024 is KB - This is just a note.
I hope this is very clear to you.
thank you,
AK G
21 REPLIES 21
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Generally, i increase the inode count. Never seen issues with performance.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
you need ot increase the inode count.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
monitor with df -i, then increase with maxfiles... you can increase but can't decrease the inode count but I haven't seen any noticeable performance issues although it is one of the warnings
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
increrased the maxfiles couple of days back..
but want to know any seen perf issues down the lane after 5-6 months later.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
I haven’t noticed any…but good to use performance advisor and perfstat if you want baseline performance information.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
thanks. will keep an eye on those volumes and see how they are performing.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
with Data ONTAP 8.0, you can decrease the maximum inode count as long as the new value is greater than the current inode allocation range.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Good point... I recently reviewed that in the release notes so any version higher than 8 can now lower the inodes which makes it a lot more flexible.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hii scott, I have a question..In my earlier company we went till 95% aggr capacity and we dint go beyond that. In my new one, they are saying we cant go beyond 85% which affects performance. Is this true..If not whats the best practice..Does it depend on disk types?.
Any suggestions would be helpful.
Thanks,
Karthik C.S
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
85% is a good practice… the issue is any full file system will not perform as well.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Thanks guys...So all our aggregates on a box reached 85%, and a new volume has to be created, I cant delete snapshots as they are important, I cant add more disks as there are only two spares left..i reduced snap reserve to 0% on aggregate already..i cant reduce inode count too as it is already 92%...Does reducing the quota limits on the qtres will provide space on aggregate as our volumes are thick provisioned…so now what am I supposed to do other than putting that volume on other storage box.
-Karthik C.S
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Changing tree quota limits will affect the user view but not the storage system used.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hi Karthik,
you don't have the options or setting to provide any space or inode's from the said the volume. Try dedup too.
Please add new storage or create new volume on another storage.
thank you,
AK G
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Yes the volume is already deduped and compressed too..So there aint any way other than adding more disks or moving volume to other storage box?
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
I don't see another way...and if you move a volume you have to wait for aggr snaps to roll off before the space will be given back...
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
cool..Thnx
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hii scott, we are planning to implement system analyzer in our environment for windows hosts. I see that it is used for ontap 7.2 or greater. So I am skeptical whether we can use it for ontap 8.1 7 mode and c-mode.
Thanks,
Karthik C.S
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hi Karthik,
The best practice is 80% -85 %, but you can increase the inodes by maxfiles parameter. I doesn't depend on the disk type. General principle every 4KB block is 1 node.
Thank you,
AK G
migration has accepted the solution
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hi Reddy,
Simple concept: Each 4K block is 1 node. lets say you have 1TB volume, the number of inodes you can have is: (1024 * 1024 *1024)/4= 268435456 inodes. The best practice is use upto 75 to 85% of nodes of a volume. If you further need the inodes, you better increase the size of the volume and increase the inodes after the threshold is reached (85% inodes consumption).
First 1024 is GB and second 1024 is mb and third 1024 is KB - This is just a note.
I hope this is very clear to you.
thank you,
AK G
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Thanks man for this information.
Do you have any NetApp document mentioning this information ?