> I noticed trying to attach two nodes with such commands: > > pcp_attach_node -d 5 127.0.0.1 9898 user pass 0 ; pcp_attach_node -d 5 > 127.0.0.1 9898 user pass 1 > > > will fail with: > > pgpool: 2010-04-14 03:44:59 ERROR: pid 4270: failover_handler: invalid > node_id -1 status:1 MAX_NUM_BACKENDS: 128 > > > On the other hand, adding sleep in between makes it succeed: > > pcp_attach_node -d 5 127.0.0.1 9898 user pass 0 ; sleep 1; pcp_attach_node -d > 5 127.0.0.1 9898 user pass 1 > > > But with "sleep", it's even less "atomic" then the first command. > > > My question: > > 1. Is it expected that it fails when we start pcp_attch_node one after > another?
No. I will look into this. > 2. Is there a way to attach multiple nodes in an "atomic" way? Why do you need atomicity for this? -- Tatsuo Ishii SRA OSS, Inc. Japan English: http://www.sraoss.co.jp/index_en.php Japanese: http://www.sraoss.co.jp _______________________________________________ Pgpool-general mailing list [email protected] http://pgfoundry.org/mailman/listinfo/pgpool-general
