Support and News Forum

Since upgrade to 5.0.1-14-PRO I have been unable to run any vMX nodes and receive the below errors once device is booted. I have also done a clean install and still get the errors. I can login as root but the 'cli' command doesn't work. Am running Eve in ESXi.

g_vfs_done():md9.uzip[READ(offset=622983168, length=2048)]error = 86
md9.uzip: UZIP(zlib) inflate() failed
g_vfs_done():md9.uzip[READ(offset=622985216, length=8192)]error = 86
md9.uzip: UZIP(zlib) inflate() failed
g_vfs_done():md9.uzip[READ(offset=46307328, length=2048)]error = 86
vm_fault: pager read error, pid 6485 (cli)
md9.uzip: UZIP(zlib) inflate() failed
g_vfs_done():md9.uzip[READ(offset=46651392, length=2048)]error = 86
vm_fault: pager read error, pid 6485 (cli)
md9.uzip: UZIP(zlib) inflate() failed
g_vfs_done():md9.uzip[READ(offset=46659584, length=2048)]error = 86
vm_fault: pager read error, pid 6485 (cli)
md9.uzip: UZIP(zlib) inflate() failed
g_vfs_done():md9.uzip[READ(offset=46675968, length=2048)]error = 86
vm_fault: pager read error, pid 6485 (cli)
md9.uzip: UZIP(zlib) inflate() failed

Help please!

and the REs of vQFXs won't start either (Junos 17.3)

Might be you can try,
wipe node/start
but anyway this lab before is running vQFX vers 20.3
EVE V5 Pro
but vMX I dorecommend to use Versions 18-21
tbh 18 was best

have tried and wiping the nodes but makes no difference. I am running 18.2 and had no issues before the upgrade. vQFX is 17.3 but that was on the advise of Christian @ Juniper as newer versions boot into the line card not master when rebooted

it's definitely an issue with the version 5. The nodes run without issue on version 4 so am going to stick with that

sidders54 wrote: ↑

Sun May 29, 2022 6:22 pm

it's definitely an issue with the version 5. The nodes run without issue on version 4 so am going to stick with that

no isssues with EVE V5 evem vMX running faster and better..

I still experiencing this issue. Seems to happen when I start a few vMXs at same time. Pain in the backside having to start them all individually and not something I had to do on previous Pro version

For my information, doesn't the delay option in the node parameter help ? Even with a short 30sec delay between each node ?