How to run distributed simulation with spectre?

[tech11]

According to the distproc.pdf, I do such setting:
1. modify /etc/hosts.equiv so that I may rlogin to other machines without
password;
2.set up the nfs server and I can read/write any data on every machine;
3.set one NTP server and make others as client to sync the time;
4.write file 'queue_config'as this:
fast 3
host1 2
host2 3
host3 5
5.edit ~/.cshrc and setenv LBS_CLUSTER_MASTER 'my_host_name' and setenv
CDS_QUEUE_CONF_FILE ~/queue_config
6.run 'su' and get root to run cdsqmgr ~harry/queue_config;
7.Open one schematic and run distributed simulation.
At the window of 'Choosing Simulator/Directory/Host', I run 'Check setup'
and get such info:
##############################################################################
Checking Distributed Processing (DP) setup..........
##############################################################################
Current User is: harry
Submission host name is: host

Checking for DP environment variables.....
##############################################################################

Environment variable LBS_CLUSTER_MASTER is set to host
Checking for LBS mode setup
Checking if cdsqmgr is running properly on "host" .........Ok
Checking if cdsNameServer is running properly on "host" .........Ok
Checking for queue config file...
Using variable CDS_QUEUE_CONF_FILE as this is set to
/export/home/harry/queue_config
********************************************************************************
The queue config file identified is /export/home/harry/queue_config
This may not be correct due to the OS command's (ps) limitations.
You can set the env variable CDS_QUEUE_CONF_FILE to specify the queue config
file used by the cdsqmgr daemon running on host
********************************************************************************
##############################################################################
Checking if the hosts are up......
host1 ....... Not up
host2 ....... Not up
host8 ....... Not up
WARNING (207): Jobs may not be submitted to the following host(s), because
these host(s) are not responding.
Consult your System Administrator to resolve the issue.
caretta13
caretta2
caretta8
##############################################################################

##############################################################################
Checking for password related requirements(from Submission to Cluster Hosts)
##############################################################################

##############################################################################
Checking for password requirements in reverse direction (from Clusters to
Submission Hosts)
##############################################################################

##############################################################################
Checking for hierarchy compatibility with login shell: none
1. Check for the hier locations
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2. Check for the presence of spectre binary.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
<<< Local spectre version : 5.0.0 >>>
3. Check for the presence of cdsserv binary.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
<<< Local cdsserv version : 06.00 >>>
##############################################################################

##############################################################################
Checking for space in /tmp area.......
##############################################################################

##############################################################################
Checking for Network Mode......
##############################################################################

##############################################################################
Checking if artmonitor works...Ok
##############################################################################


Setup check completes with 0 errors and 1 warnings.


I don't know if I miss something, actually when I run adptest, I get such
info:

$ adptest

BEGIN Artist Distributed Processing Setup Test

Testing for existence of LBS_CLUSTER_MASTER variable...
PASSED.

Testing opening of Artist Job Server...
Launching Artist job server...
Artist job server has initialized!
PASSED.

Please be patient - the remaining portion of the test takes some time.

Testing job submission to host `host1' on queue `fast'...
PASSED.
*Error* MPS: bad Handle - mpsBAD:0x0x8202018
*Error* load: error while loading file -
"/export/cdsdir/tools/dfII/etc/tools/adp/adptest_enc.ils"

I can visit file "/export/cdsdir/tools/dfII/etc/tools/adp/adptest_enc.ils"
also.

I have no more idea to this. Who may help me or give me some advice? Thanks
for your help!

Have a good day!

Best Regards,

Harry

 

[JoRobins]

I saw a whole bunch of suggestions in one of the spectre FAQ docs on
sourcelink on how to debug distributed sim problems. I found those to
be quite helpful.. Have you checked that and tried out the suggestions?
Some that I remember
- did you try killing the cdsnamemgr process as well before restarting
the cdsqmgr?

Also I got an error while I did the "check setup" utility (something
about being unable to identify the OS). But when I really ran the
simulation from the Artist menu, the distributed sim ran without any
problem. Not sure if you tried that..