Discussions
Categories
Groups
Community Home
Categories
INTERNAL ENABLEMENT
POPULAR
PUBLIC CLOUD
PRIVATE CLOUD
Quick Links
MY LINKS
HELPFUL TIPS
Back to website
Home
Web CMS (TeamSite)
SAN and Performance
Scrubbs
Looking to Install 6.5 in the near future and the Company is going to SAN. Is anybody else using SAN.
Have you seen any performance issues? Has performance been better?
Find more posts tagged with
Comments
Bill Klish
We are using a Network Appliance Filer to contain our backing stores and have not noticed any performance degradation. The downtime to freeze TeamSite and do the backup is less than 2 minutes for our 15 GB backing stores. We utilize the snapshot capability to perform the work.
Our TeamSite servers are solaris servers that have dual network cards. One that handles all incoming requests and another gigabit card on a gigabit internal network ring to handle all requests to the filer.
Interwoven has sent out a whitepaper about this and could send it to you if you wish. I don't think it is posted in any KB article or the library. Let me know.
Johnny
I'd be interested in seeing something on this subject.
John Cuiuli
Migrateduser
If you get me this whitepaper I can research where it's IWIOV location is and let people know.
lissa
psr
Hi bk99nd,
We are also planning to implement a simillar solution. I will appreciate if you could fwd me the PDF.
!Thanks
Bill Klish
Sorry for the delay. I will forward the document to you so you can post it in the proper place. It is fairly large.
Migrateduser
Hi,
I am also interested in the paper. Can you send me a copy, too?
Thanks,
Migrateduser
I'm checking with the folks who produced this paper to make sure it is current. I'll be in touch.
lissa
Scrubbs
We have talked with Interwoven they have said no performance problems with SAN. In talking with our IT
people also we are confident about SAN. Looking at setting up a development area using SAN with Teamsite 6.5
to ensure no problems. Would also like to see the document being talked about.
rpatiban
I wonder if anybody has installed TeamSite itself on a SAN drive.
I would appreciate any inputs in this regard,
Thanks much in advance,
Raghu
kish2
can i know where this doc is - we are about to implement SAN too?
Thanks
Migrateduser
I asked some folks to review the doc. it was over 3 years old, and I wanted to be sure it was still ok. I will ping them again.
But I don't know of any issues with SAN.
Regards,
lissa
Bill Klish
Yes. We have had TeamSite 5.5.2 and 6.1 SP1 (the backing store) on a network appliance filer for over 2 years now. The downtime to backup is around 1 minute to take a snapshot of the volume. We see no performance degradation at all, and have trimmed our backup window from 2 hours to 1 minute.
Cust_Orders_Payments.rptdesign
Evansw
Hi,
We currently have our live TS environment backing store on a SAN and would like to be able to copy this backing store across the network to another SAN within a different environment.
Does this document you mention have some information to do this? We have tried Rsynch but it took a while!.
We are on solaris O/S, running 5.5.2 in live environment and 6.5 in copy environment, and our backing store is around 13Gb? Is there an easy way of getting this data copied across without it taking 15+ hours?
Please help!
Adam Stoller
Since you apparently need to *copy* 13Gb from one server/disk to another server/disk - you need to account for (a) network latency and (b) disk reading/writing latency.
You *might* be able to reduce time on [subsequent] updates by doing something like this:
1. Copy (rsync, OD, whatever) from SAN-1/location-1 to SAN-2/location-1
2. Copy (rsync, OD, whatever) from SAN-2/location-1 to SAN-2/location-2
SAN-1/location-1 would be the 5.x backing store
SAN-2/location-1 would be sort of a staging area for your copy - representing the last known copy from SAN-1/location-1
SAN-2/location-2 would be the working backing store locaton for your 6.5 server.
The initial copy [1] will still take a long time.
By using a comparitive copying utility (rsync, OD, whatever) for [1], all subsequent copies [1] will in effect be incremental updates and should not take as long.
By using a comparitive copying utility for [2] and having it act between two locations (probably partitions) on the same server - removes the network latency issue from the equation and should therefore be generally faster than if you periodically copied from SAN-1/location1 to SAN-2/location-2
You would probably be best off if you added another indirection to the equation (S=SAN, L=location):
(a) S-1/L-1 => S-1/L-2
(b) S-1/L-2 => network => S-2/L-1
(c) S-2/L-1 => S-2/L-2
(a) requires that you stop the 5.x TS server temporarilly to create a backup (if your SAN provides this kind of functionality itself - all the better)
(b) then works off of essentially a read-only copy from S-1 to a staging location on S-2
(c) requires that you stop the 6.x TS server temporarilly to re-synch the data in L-2 from L-1
--fish
Senior Consultant, Quotient Inc.
http://www.quotient-inc.com
Bhargav Coca
I aminterested in reading this documentation what everyone was talking about, can you post it some where so that everyone can access it, atleast people with support prevelages
Thank you
Migrateduser
I've still got the whitepaper. It was VERY old. I asked the engin team to review it, they wanted to make a few changes, and I haven't gotten it back.
I'll ping them again.
lissa
Migrateduser
I was able to get the whitepaper reviewed. It is up on DevNet now in the techlib:
http://devnet.interwoven.com/site.fcgi/techlib/docs/site.fcgi/techlib/54529
hth,
lissa