09-13-2019, 12:16 AM
(09-12-2019, 10:41 PM)mark_talluto Wrote: Hi Simon,
I modified the firewall to allow ping to work on your region. Please give cdb_pingNode() another try.
The variability in transaction time when doing a cdb_ping() call could be due to the activity level of a given instance. cdbPing() is useful to gauge how available the portion of the region your data is stored. cdb_pingNode() is useful to guage how fast you can get a response from your portion of the region under the best conditions.
Thanks Mark,
New results with cdb_pingNode ()
Quote:cdb_pingNode(): 132.501
cdb_ping(Notifications): 824.40114
---------
cdb_pingNode(): 126.351
cdb_ping(Notifications): 589.275837
---------
cdb_pingNode(): 142.869
cdb_ping(Notifications): 565.644979
---------
cdb_pingNode(): 128.699
cdb_ping(Notifications): 587.082863
---------
cdb_pingNode(): 127.671
cdb_ping(Notifications): 615.654945
---------
Total elapsed time: 4166
and with cdb_pingNode ("Notifications")
Quote:cdb_pingNode(Notifications): false
cdb_ping(Notifications): 1172.601938
---------
cdb_pingNode(Notifications): false
cdb_ping(Notifications): 966.895103
---------
cdb_pingNode(Notifications): false
cdb_ping(Notifications): 936.855078
---------
cdb_pingNode(Notifications): false
cdb_ping(Notifications): 962.046862
---------
cdb_pingNode(Notifications): false
cdb_ping(Notifications): 942.969084
---------
Total elapsed time: 5146
Still getting "false" with cdb_pingNode("Notifications"), and huge variations on cdb_ping("Notifications")
Bottom line: what would you suggest about how to check for data availability before shooting a transaction against such data or sending an error message to the user asking to try again later.
To ";" or not to ";" that is the question