Azure Virtual Network
An Azure networking service that is used to provision private networks and optionally to connect to on-premises datacenters.
2,778 questions
Increased Latency - On-Premise servers vs. Azure Servers
![](data:image/svg+xml, %3Csvg xmlns='http://www.w3.org/2000/svg' height='64' class='font-weight-bold' style='font: 600 30.11764705882353px "SegoeUI", Arial' width='64'%3E%3Ccircle fill='hsl(108.80000000000001, 65%, 20%)' cx='32' cy='32' r='32' /%3E%3Ctext x='50%25' y='55%25' dominant-baseline='middle' text-anchor='middle' fill='%23FFF' %3EAL%3C/text%3E%3C/svg%3E)
AJ Ludenstein
0
Reputation points
Hello All-
Would assigning the Citrix Server subscription to our Application Server subscription reduce latency?
Server Subscription Name Example
Ctx Server: IT Prod Hosting
App Server: Engineering Prod Hosting
Scenario:
-We've migrated our Dev, QA, and Prod Citrix and Application servers to Azure
-Citrix is used to access the application required for production data
-Citrix uses a different subscription from the application servers
-when operators login to Citrix and then login to the application latency has significantly increased
-latency has increased from approx. < 15 seconds (on-premise servers) to > 1minute 15 seconds (Azure servers)
Azure Virtual Network
{count} votes
-
KapilAnanth-MSFT 49,616 Reputation points Microsoft Employee Moderator2023-05-23T04:58:27.22+00:00 Welcome to the Microsoft Q&A Platform. Thank you for reaching out & I hope you are doing well.
Per my understanding,
- You have a CITRIX server in subscriptionA
- Application server in subscriptionB
- Users login to Citrix and from there, they login to Application server
- And the latency seems to have increased post migration to cloud.
Please correct me if I am wrong.
1.Would assigning the Citrix Server subscription to our Application Server subscription reduce latency?
- No.
- Subscription is merely a logical grouping for Billing resources.
- The factors that would influence latency are region they are hosted and the method in which these two servers are connected to each other.
You can consider using Proximity placement groups
- While deploying the VMs in same subscription will not improve the latency, using a PPG will greatly improve latency.
- NOTE : For using a PPG, the VMs should be in same subscription
Now, wrt your scenario,
- Are the CITRIX server and Application server hosted in same region or different regions?
- How are these two servers connected?
- Are you using VPN or VNET Peering or just Internet to establish connectivity?
- Will you be able to directly ping/telnet the Application server from CITRIX Server?
- How are you calculating the latency here?
Cheers,
Kapil
-
AJ Ludenstein 0 Reputation points
2023-05-23T22:08:12.8133333+00:00 Yes you are correct. Thank you for your quick response!
- Are the CITRIX server and Application server hosted in same region or different regions?
-> Citrix - US VA Azure East Subscription A
-> App - US IL Azure North Central Subscription B
-> SQL DB -
- How are these two servers connected?
- Are you using VPN or VNET Peering or just Internet to establish connectivity? -> VNET
- How are you calculating the latency here?
-> manually timed
- How are you calculating the latency here?
- Are you using VPN or VNET Peering or just Internet to establish connectivity? -> VNET
Additionally, the firewall for subscription B is more locked down than subscription A.
-
KapilAnanth-MSFT 49,616 Reputation points Microsoft Employee Moderator
2023-05-24T03:52:52.96+00:00 Thanks for the update.
So,
- The CITRIX server is in East US and Application server is in North Central US
- And you are using VNET Peering and testing latency manually.
Here is a Region-to-Region latency chart for Azure backbone :
- Refer : Region to Region Latency
- From this table, we can see that it should be 20 ms.
- So, I would say the latency you are seeing is expected behavior.
To improve performance and decrease latency,
- You should consider deploying the VMs in the same region.
- Even in same region, to further decrease latency, consider using Proximity placement groups
If you'd like to Troubleshoot the current scenario,
- Please share a ping latency test between the servers
- From CITRIX to Application
- And from Application to CITRIX
- Do a TCP Ping such as psping and share the results
- From CITRIX to Application
- And from Application to CITRIX
- Do a tracert
- From CITRIX to Application
- And from Application to CITRIX
- From CITRIX to Application
- When you say the Application subscription is more locked down,
- Is there any other intermediate hop between CITRIX and Application?
- Any firewalls or NVAs?
- Please share a network diagram of your architecture
- Ideally, I am hoping the connectivity should be direct like below
- CITRIX server ------------ Global VNet Peering -------- Application
- If this is not the case, please elaborate.
Cheers,
Kapil
-
KapilAnanth-MSFT 49,616 Reputation points Microsoft Employee Moderator
2023-05-25T08:34:46.8533333+00:00 Can you please update us if the action plan provided was helpful?
Should there be any follow-up questions or concerns, please let us know and we shall try to address them.
Thanks,
Kapil
-
AJ Ludenstein 0 Reputation points
2023-05-25T18:31:00.1366667+00:00 My apologies I've made a mistake with the connections.
Subscription A uses a connection called express route
Subscription B uses VPN
-
![](data:image/svg+xml, %3Csvg xmlns='http://www.w3.org/2000/svg' height='64' class='font-weight-bold' style='font: 600 30.11764705882353px "SegoeUI", Arial' width='64'%3E%3Ccircle fill='hsl(115.19999999999999, 24%, 21%)' cx='32' cy='32' r='32' /%3E%3Ctext x='50%25' y='55%25' dominant-baseline='middle' text-anchor='middle' fill='%23FFF' %3EM%3C/text%3E%3C/svg%3E)
michaelanna 0 Reputation points2023-05-25T18:36:25.0233333+00:00 When comparing on-premise servers to Azure servers in terms of latency, several factors come into play that can impact the overall latency experienced. Here are a few key points to consider:
- Proximity to Users: With on-premise servers, the physical distance between the server location and the users can significantly affect latency. If the users check are geographically dispersed and far from the on-premise server, it can result in higher latency compared to Azure servers, which have a global presence and can provide servers closer to users in different regions
-
KapilAnanth-MSFT 49,616 Reputation points Microsoft Employee Moderator
2023-05-26T04:40:06.77+00:00 Thanks for the info
That actually indicates there's more to your architecture than we are seeing.
Can you please share a Hop to Hop network architecture of your environment?
And also a tracert result
- The statement "Subscription A uses a connection called express route.Subscription B uses VPN" is incomplete.
- This only indicates there's more components in between the VPN and ExpressRoute
In this scenario,
- I would say it's best to deploy them in the same subscription and region (if not, atleast same region - this will drastically improve performance)
- And Connect the VNets where the VMs are deployed using VNet Peering only
- Implementing **Proximity placement groups **with the above should help further
Cheers,
Kapil
-
KapilAnanth-MSFT 49,616 Reputation points Microsoft Employee Moderator
2023-05-29T05:04:50.84+00:00 Reaching out to check if there are any questions on this.
Please let us know if we can be of any further assistance here.
Thanks,
Kapil
-
KapilAnanth-MSFT 49,616 Reputation points Microsoft Employee Moderator
2023-05-31T08:11:15.28+00:00 May I know if you got a chance to review my previous comment?
Please let me know if you are facing any challenges or if there are any follow-up questions, I shall be glad to address them.
Thanks,
Kapil
Sign in to comment
Sign in to answer