Here are my numbers for PROTECT STG
        
Server pair 1 (2*10 Gb/s link, 25km, TCPWINDOWSIZE 0,  default #threads):       
Max: 381 Mbytes/s    -  Normal 100-200 Mbytes/s
Server pair 2 (2*10 Gb/s link, 25km, TCPWINDOWSIZE 0,  default #threads):       
Max: 242 Mbytes/s    -  Normal 100-200 Mbytes/s
Server pair 3 (2*10 Gb/s link,  1km, TCPWINDOWSIZE 0,  default #threads):       
Max: 210 Mbytes/s    -  Normal 100 Mbytes/s
Server pair 4 (2*10 Gb/s link,  1km, TCPWINDOWSIZE 0,  default #threads):       
Max: 197 Mbytes/s    -  Normal 100-150 Mbytes/s 

All number a BYTES not bits per second.   Protect STG runs every hour.

Best Regards
Anders Räntilä

-----Original Message-----
From: ADSM: Dist Stor Manager [mailto:ADSM-L@VM.MARIST.EDU] On Behalf Of 
Matthew McGeary
Sent: den 1 december 2016 21:15
To: ADSM-L@VM.MARIST.EDU
Subject: Re: [ADSM-L] Storage container replication

David,

We're able to saturate our 1Gbps link to our DR site during stgpool protect 
processes.

The trick is to specify a relatively high number of sessions (50 in this case) 
to stick-handle around the latency.  We've also enabled the sliding TCP window 
on both source and target and raised the buffers high enough for the servers to 
allocate an 8M window size.

Our before-dedup intake is around 15TB a day, which translates to around 1-2TB. 
 With two protect processes scheduled to run at 1:30 am and pm, they tend to 
run to completion in 2-3 hours.

HTH
__________________________
Matthew McGeary
Senior Technical Specialist - Infrastructure Management Services PotashCorp
T: (306) 933-8921
www.potashcorp.com


-----Original Message-----
From: ADSM: Dist Stor Manager [mailto:ADSM-L@VM.MARIST.EDU] On Behalf Of David 
Ehresman
Sent: Thursday, December 01, 2016 1:03 PM
To: ADSM-L@VM.MARIST.EDU
Subject: [ADSM-L] Storage container replication

If you are doing directory based storage containers and using "protect stg" to 
replicate data from a primary site to a secondary site, what kind of throughput 
are you getting on the data movement?

David

Reply via email to