windows 10 host fedora 29/centos 7 guest bridged networking fails under vmware workstation

[mtk]

after a recent windows 10 update (to v1809, on or around mar 8/9), bridged networking fails for both fedora 29 and centos 7.  in both cases, the interfaces are marked up and the standard routes are present.  but any attempt to ping something on the subnet or through the router fails with "host unreachable".   i can ping another VM running within the same vmware process (e.g. F29 to RHEL7 or vice versa).  i've restored the network to the default outside of vmware with the virtual network app.  but it had no affect.  thoughts/suggestions? 

if i try to 'report a bug' from the vmware workstation 'support menu', i'm told that support has expired.  this seems counter productive since it should be your benefit to hear about this.  i'd also like search results to be limited to a particular (recent) product release.

/mark

[Speed3]

I am having the same issue with Windows 10 v1809 host running Windows 7 on Workstation 15 Pro. The bridged network does not work. Tried all the recommended fixes I've read and nothing has helped.

Todd

[mtk]

i'm windows 10 up-to-date as of the patches released yesterday, 3/12/19.  i uninstalled vmware workstation pro 15 and reinstalled from scratch.  no effect (it was the same software version, after all).

clearly, since it affected both a fedora 29 VM and a centos v7 VM, this problem has nothing to do with the OS's themselves (i.e. unlikely that it is vmtools-dependent).  given that the recent windows updates have focused on spectre/metldown/retpoline, i'm guessing some big glitch was introduced at a low level.

i can't submit a bug report since i'm long out of the one month support period.  i've been a vmware workstation customer going back to at least v6.  i feel like i shouldn't have to pay $50 to file my bug report.

vmware, are you listening?

/mark

[mtk]

there are patches for at least vsphere.  really need your response here, vmware.

[mtk]

should also add that before bridged networking failed completely for fedora 29 and centos 7, fedora 29 booting always produced a dark screen instead of the sign on screen.  if i paused and resume the VM, the real login screen would display normally.  shutdown of fedora 29 also started to hang when invoked from the OS.   shutdown from the workstation menu works.  no such problems with centos 7.

[mtk · Accepted solution]

just changed the networked setting from automatic bridging to bridging via the physical network interface and now everything is working!  (a suggestion from another thread).  the automatic bridging setting is what you get if you reset to defaults.  one would think that is correct.  can anyone explain this?

/mark