Thank you for the follow up information.
I was thinking to add another reply, asking whether the issue was only seen with one particular host, and if so, for troubleshooting, often it's best to initially focus just on that host (for problems unique to that host, such as you initially noted and with the follow up).
If a performance issue is more widespread, like many or all hosts, whose traffic transits a particular network device or link, that points more to a network issue.
But, even with "hints", like the forgoing, some issues can be a combination of the network device and specific hosts, such as flow control, because adjacent devices both must have it enabled to be actively used (and a adjacent device can be a host). As device device defaults can differ, it can sometimes be unclear that it takes a combination of devices, with specific default configurations, to cause a problem, and, don't forget, something like flow control might be ideal for some neighboring device combinations, so, again, it's easy to miss the cause of some issues.
Also again, glad to read, it seems, the problem its just on a single host.
Laugh, often my network troubleshooting, has been used to show the network isn't the problem, but where the problem looks to be. Much easier to just say, the problem is due to "x" (like sun spots, laugh) rather than do the work to run down the problem.
Networking engineers, often get used to the network being blamed for all performance problems, when it's so often not, they get used to, more or less, "wasting" their time, to "prove" it's not a network issue, once again.
But, sometimes, it is non obvious network problem. (My favorite story, I like to tell, we had a MetroE hand-off to a tier-one service provider. I noticed, and complained to the SP, I don't believe it's working quite right, as I seem to be losing some packets across your cloud. They checked, doubled checked and tripled checked, it's working just fine, there are no drops, you just don't understand how it's supposed to work [as the physical hand-off had more bandwidth than the CIR]. I stated, oh, I think I do understand, and still believe it's not quite right. So, since I kept complaining, they went to my networking manager, and complained I kept complaining. My network manager backed me, if Joe says it's not quite right, it probably isn't. After 3 months, they finally discovered out-of-date firmware, on one of the line cards, being used, which had a minor bug, when utilization would surpass 97%, there were some/few additional, needless, drops, also not recorded on their interface. They updated the firmware [surprise] now it worked as I expected.)
