replicator node dies every time

[Jailander]

Hi,

I am testing the released version of mongodb_store and launching it always results in the replicator node crashing, this doesn't happen to me with the version on my strands_ws however it may still be the same issue as #52 since running rosrun mongodb_store replicator_node.py after the datacentre has been launched works without any problem and the log is the same:

Traceback (most recent call last):
  File "/opt/ros/hydro/lib/mongodb_store/replicator_node.py", line 141, in <module>
    store = Replicator()
  File "/opt/ros/hydro/lib/mongodb_store/replicator_node.py", line 29, in __init__
    raise Exception("No master datacentre found using mongodb_host and mongodb_port")
Exception: No master datacentre found using mongodb_host and mongodb_port
[INFO] [WallTime: 1415029899.597475] Mon Nov  3 15:51:39 [initandlisten] waiting for connections on port 62345
[INFO] [WallTime: 1415029899.597658] Mon Nov  3 15:51:39 [websvr] admin web console waiting for connections on port 63345
[INFO] [WallTime: 1415029899.629664] Mon Nov  3 15:51:39 [initandlisten] connection accepted from 127.0.0.1:34111 #1
[INFO] [WallTime: 1415029899.630524] Querying content in a futher 0 datacentres
[INFO] [WallTime: 1415029899.668076] Mon Nov  3 15:51:39 [initandlisten] connection accepted from 127.0.0.1:34119 #2
[INFO] [WallTime: 1415029899.669733] Default parameters path not supplied, assuming none.
[replicator_node-4] process has died [pid 20155, exit code 1, cmd /opt/ros/hydro/lib/mongodb_store/replicator_node.py __name:=replicator_node __log:=/home/jaime/.ros/log/ef5a5470-6370-11e4-8c52-902b34d5fb2e/replicator_node-4.log].
log file: /home/jaime/.ros/log/ef5a5470-6370-11e4-8c52-902b34d5fb2e/replicator_node-4*.log
[INFO] [WallTime: 1415029959.598790] Mon Nov  3 15:52:39 [clientcursormon] mem (MB) res:31 virt:920 mapped:80

[marc-hanheide]

looks pretty much like the same problem #52. One cannot assume any order in launch files. So this requires active waiting and checking for some period until reaching a timeout I'd say. I'll mark this as a duplicate of #52 and assign @hawesie to it to check if that is the problem. Then he shall assign it to someone to fix it :-)

[hawesie]

Closing as a duplicate of #52.