VM Does not work in VM Player

[kovacsbv]

I tried to open HPCCSystemsVM-3.0.3.3.vmx, and when the machine started up, I got:
00000000 2011-08-04 12:51:02 1041 "jsocket(1,2778) getaddrinfo failed err = -3 : localhost.localdom"
/opt/HPCCSystems/etc/init.s/hpcc_common: eval: line 69: syntax error near unexpected token `)'

The error repeats on line 121
Looking into the script, there are some quotes that need to be escaped if you want to run the script like that:

Code: Select all

eval "$(echo "${cmp[*]}")"

Should be more like:

Code: Select all

eval "$(echo \"${cmp[*]}\")"

Then I get the vami-sfcbd restarting 10 times (slowly).

In the end, netstat -pln reveals that port 8010 is not open, although 3 548x ports are open for IPv6 only.

Code: Select all

Versions:
OS:      XP 32
HPCC:    3.0.3.3
VMWare:  Player 3.1.4

Thanks,

Vic Kovacs

[sort]

Thank you for pointing this out to us.

We have made some code changes in our platform that should make the "getAddrInfo" error no longer occur. Certain ports like 8010, 8002 need to be open

We have not made any changes yet to the script in our most recent release that address the syntax errors. We will look to correct our problem across all the distros we support. We appreciate the code you provided in your post

We will be releasing our source code in the near future and look forward to you submitting code to us.

Thank you
Stu

[kovacsbv]

Ok, This error was seen before on VMWare's forum at http://communities.vmware.com/message/1712897?tstart=0.

This page refers to the vami-sfcb log to look for errors:

I found the errors in /opt/vmware/var/log/vami/vami-sfcb.log

Code: Select all

05/08/2011 15:35:11 [info] [process id: 2526] Starting a new initialization script.
05/08/2011 15:35:11 [info] [process id: 2526] Pids passed in are 2523 2522 2521 2519. Sleeping for 300 (secs).
05/08/2011 15:40:11 [info] [process id: 2526] ran command: ps -p 2523 2522 2521 2519
05/08/2011 15:40:11 [info] [process id: 2526] ps returned process:   PID TTY      STAT   TIME COMMAND
05/08/2011 15:40:11 [info] [process id: 2526] ps returned process:  2519 ?        Ssl    0:00 /opt/vmware/sbin/vami-sfcbd -d
05/08/2011 15:40:11 [info] [process id: 2526] ps returned process:  2521 ?        S      0:00 /opt/vmware/sbin/vami-sfcbd -d
05/08/2011 15:40:11 [info] [process id: 2526] ps returned process:  2522 ?        S      0:00 /opt/vmware/sbin/vami-sfcbd -d
05/08/2011 15:40:11 [info] [process id: 2526] ps returned process:  2523 ?        S      0:00 /opt/vmware/sbin/vami-sfcbd -d
05/08/2011 15:40:11 [info] [process id: 2526] Number of sfcbd processes: 4
05/08/2011 15:40:11 [info] [process id: 2526] VAMI Scheduler provider is present on the system.
05/08/2011 15:40:11 [ERROR] [process id: 2526] Failed to start Scheduler Service. Unknown error.
05/08/2011 15:40:11 [ERROR] [process id: 2526] Traceback (most recent call last):
  File "/opt/vmware/share/vami/vami_sfcb_initialize", line 86, in main
    rettuple = vami_cim_util.invokeNonStaticMethod('root/vami', 'VAMI_SchedulerService', 'StartService')
  File "/opt/vmware/share/vami/vami_cim_util.py", line 36, in invokeNonStaticMethod
    insts = cliconn.EnumerateInstanceNames(className)
  File "/opt/vmware/lib/python/site-packages/pywbem/cim_operations.py", line 382, in EnumerateInstanceNames
    **params)
  File "/opt/vmware/lib/python/site-packages/pywbem/cim_operations.py", line 173, in imethodcall
    raise CIMError(0, str(arg))
CIMError: (0, 'Socket error: [Errno -3] Temporary failure in name resolution')
05/08/2011 16:19:56 [info] [process id: 2463] Starting a new initialization script.
05/08/2011 16:19:56 [info] [process id: 2463] Pids passed in are 2460 2459 2458 2456. Sleeping for 300 (secs).
05/08/2011 16:24:56 [info] [process id: 2463] ran command: ps -p 2460 2459 2458 2456
05/08/2011 16:24:56 [info] [process id: 2463] ps returned process:   PID TTY      STAT   TIME COMMAND
05/08/2011 16:24:56 [info] [process id: 2463] ps returned process:  2456 ?        Ssl    0:00 /opt/vmware/sbin/vami-sfcbd -d
05/08/2011 16:24:56 [info] [process id: 2463] ps returned process:  2458 ?        S      0:00 /opt/vmware/sbin/vami-sfcbd -d
05/08/2011 16:24:56 [info] [process id: 2463] ps returned process:  2459 ?        S      0:00 /opt/vmware/sbin/vami-sfcbd -d
05/08/2011 16:24:56 [info] [process id: 2463] ps returned process:  2460 ?        S      0:00 /opt/vmware/sbin/vami-sfcbd -d


[kovacsbv]

The errors are beginning to confirm something I've been suspecting: The VM wants a FQDN that DNS can resolve into an IP. Maybe I can set up a hosts entry. I'll look for alternatives, but it looks a bit like I might need a static IP with a DNS entry.

Right now, it looks like initialization reads /etc/hostname and uses that to set the host name which it then tries to look up with DNS, and can't.

I'll play with /etc/resolv.conf and /etc/host* and see if it can be fixed.

Of course, this may be completely unrelated to why port 8010 won't open up. Can anybody tell me which executable opens the port and listens, and where its log file is?

[richardkchapman]

port 8010 will be opened by esp executable, log file will be in /var/log/HPCCSystems/myesp/esp.log

[pschwartz]

Vic,

Thank you for taking the time to alert us of the issue. In the release on 08/05 a few of the issues you are reporting have been addressed. Please try the 3.0.4 release at your convenience.

The main issue at hand is the lack of an ip address being assigned on the first boot of the VM on your instance of VMware Player. I have personally seen this issue before and here are a few things I have done to correct it, one of which might help you.

1. Verify that the VMWare vnet device is functional.
2. Reboot your host machine to allow the VMWare vnet driver to reload.
3. Verify that the windows firewall is allowing network communications for VMWare Player.
4. Reinstall VMWare Player.

vami-sfcbd is a vmware component that is set to use the localhost as its resolving address which if the ip address is not received correctly is not set in /etc/hosts causing the component to fail to start and sometimes become stuck in an infinite restart loop. I have seen a few open trouble tickets on the VMWare support site that indicate that this issue is already being looked at by VMWare and will be corrected in future versions of their software.

As for the issue of port 8010 not being accessible, the usage of 3.0.4 should resolve this issue for you.

[kovacsbv]

Very nice.
That version cleaned it all up.
Thanks for your help.

Vic Kovacs