- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
File copy via SMB is slow over site to site vpn
Hi Guys
Need your expert advise. I have a site to site VPN via spoke-hub setup.
Ping and traceroute results are optimal at less than 50ms from both sites. Internet works fine and fast but once it comes to accessing the file server via usual unc path or copying to & fro the files, the speed only hits 300kbps or even goes to 0 then spike up n stuck at 300kbps.
Did a pcap and saw re transmissions of [tcp dup ack] messages.
Anyone encounter such issue? strangely only this site having this issue. its lan(spoke) is all connected to the switch 2960 on 100Mbps while the hub where the destination file server is a a gigabit based environment.
Thanks for the help in advance
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Are there any other issues in the site or is it just the site to site vpn?
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Tried turning off the AMP and IDS didnt help much
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
The reason I asked if you have any other problems is because a lot of DUP ACK's could be caused by a bad (physical) link somewhere in the chain. A bad wire, or a cable that's too long, crosstalk, ... But then I would expect you to see other problems on the site too, not just the site-to-site VPN. I'd try to test another protocol just to see if it's SMB specific.
Also, what bandwidth are the WAN links on both ends?
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
I have seen this issue before where the circuit is asymmetric, specifically it has asymmetric timing. For example, a packet one way takes 10ms and the return packet takes 40ms.
If the server and client are running a modern Windows TCP stack (Windows 7 or better) try running this command on both the server and the client, and repeat the test:
netsh int tcp set global timestamps=enabled
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Same exact problem for me, ever find a solution?
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Did you ever get a resolution to this
I have similar problem
Hub and spoke on 18.107.10
Latency 34ms
Hub 1Gbps symmetric circuit
Spoke 300Mbps symmetric circuit
Internet access from Spoke via Hub with central breakout
Internet access good speeds and upload/downloads to Onedrive good
Windows SMB (V3) file shares upload from spoke to hub good
downloads from Hub to spoke very slow
All other applications appear good.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
I've determined this appears to be unrelated to meraki in my case, the expressroute connection to the Azure datacenter is somehow at fault. When using a different azure datacenter with a VMX inside for site to site, there are no more smb issues.
