(This message has been automatically imported from the retired mailing list)
Minor bug report. When you start the load balancer, if there’s an error
due to existing pidfile it gives a zero result code; this means that the
zato-qs-start.sh script continues happily as if everything were OK, when
it isn’t.
To replicate:
zato@cfprov:~/env/qs-1$ BASE_DIR="$( cd -P “$( dirname “$SOURCE” )” &&
pwd )"
zato@cfprov:~/env/qs-1$ ZATO_BIN=zato
zato@cfprov:~/env/qs-1$ cd $BASE_DIR/load-balancer
zato@cfprov:~/env/qs-1/load-balancer$ $ZATO_BIN start .
Error - found pidfile /opt/zato/env/qs-1/load-balancer/zato-lb-agent.pid
zato@cfprov:~/env/qs-1/load-balancer$ echo $?
0
zato@cfprov:~/env/qs-1/load-balancer$ telnet localhost 11223
Trying ::1…
Trying 127.0.0.1…
telnet: Unable to connect to remote host: Connection refused
However in a similar situation, the server daemons do correctly report
the error.
Regards,
Brian.
Hello Brian!
Thanks for reporting it, I have just started building
2.0.2 which will include a change to cover it.
Regards!
On 25.02.2015 15:37, Brian Candler wrote:
Minor bug report. When you start the load balancer, if there’s an error
due to existing pidfile it gives a zero result code; this means that the
zato-qs-start.sh script continues happily as if everything were OK, when
it isn’t.
To replicate:
zato@cfprov:~/env/qs-1$ BASE_DIR="$( cd -P “$( dirname “$SOURCE” )” &&
pwd )"
zato@cfprov:~/env/qs-1$ ZATO_BIN=zato
zato@cfprov:~/env/qs-1$ cd $BASE_DIR/load-balancer
zato@cfprov:~/env/qs-1/load-balancer$ $ZATO_BIN start .
Error - found pidfile /opt/zato/env/qs-1/load-balancer/zato-lb-agent.pid
zato@cfprov:~/env/qs-1/load-balancer$ echo $?
0
zato@cfprov:~/env/qs-1/load-balancer$ telnet localhost 11223
Trying ::1…
Trying 127.0.0.1…
telnet: Unable to connect to remote host: Connection refused
However in a similar situation, the server daemons do correctly report
the error.
Regards,
Brian.