Link to home
Create AccountLog in
Avatar of canuckconsulting
canuckconsultingFlag for United Kingdom of Great Britain and Northern Ireland

asked on

How to validate health of network connecting between servers located in office and those on cloud

We are having performance issues regarding data replicating between a DB server located on our local server and another on the cloud.

I want to check that the network between the nodes if functioning as expected over the  2mb dedicated pipe we have.  I need some guidance on tools to use and how to interpret results to assess the health and performance of the network.

I've started and included trace routes done from both servers shown below.  Perhaps these are enough though I'm uncertain if the times are what I should be expecting.

From Local to Cloud:

User generated image
From Cloud to Local:

User generated image
Avatar of rharland2009
rharland2009

2mb pipe isn't exactly huge. What is the physical distance in miles between your premise and your cloud location?
Avatar of canuckconsulting

ASKER

The office is based in Florida and I believe the cloud location is in Maryland so say 1000 miles roughly.
Okay. Given the distance and the assumption that the connection right now is not under excessive load, then 45msec isn't out of the realm of ordinary latency.
It might be interesting to perform these same tests while attempting replication.
What is the size of the DB/dataset you're attempting to replicate across this pipe?
I'll do another test when replication kicks off tomorrow morning and post the results.

The size of the entire db (mdf file) is just under 2GB.
SOLUTION
Avatar of rharland2009
rharland2009

Link to home
membership
Create an account to see this answer
Signing up is free. No credit card required.
Create Account
I'm doing snapshot replications.  The reason for this is every morning this the source database on local is completely refreshed from a non-sql data source.  I then take a snapshot which replicates over to the cloud.  Since all the tables on the local db are truncated I thought snapshot was the way  to go.  Does this approach make sense to you?
It's probably the safest way to make sure you have a complete dataset in the cloud post-replication, yeah. Bandwidth limitations are always a challenge.
Just to verify, when you say 'performance issues', do you mean the replication takes a long time, or fails, or speed to the net stinks during replication?
It seems to slow at points.  When I look in Replication Monitor it sometimes waits for some time on "Waiting for System response". I was wondering if that might have been caused by network faults or some other related limitation.

I guess what I need to do is work out exactly how much data is transferred (not just look at the size of the mdf) and how long it takes.  I can then compare actual mb/sec and compare that to the ideal.  Is there a way to do that?  I can tell from Replication Monitor how long it took but I'm not sure how to work out how many mb were transferred.
ASKER CERTIFIED SOLUTION
Link to home
membership
Create an account to see this answer
Signing up is free. No credit card required.
Create Account
Thank you!
It's a pleasure to help.