Cluster usage and remote MySQLDB

2 posts / 0 new
Last post
KC0VKN
Cluster usage and remote MySQLDB

Recently I've been using cqrlog on several systems; one right next to the system hosting the mysqlDB, and a couple others that are far remote and accessing over an ssh tunnel.
 
I've noticed that if I have the cluster active, very frequently I'll recieve the mysqlDB error message where the remote is unreachable (though, it still is, as there is a local mysql monotor window running over the same path..).
 
This will happen with either local/remote instances of cqrlog and it's really much much more frequent if I've got the cluster on.  I suppose there could be enough cluster activity as to cause the msql-connection verifier to miss an attempt; how many times must the mysql command fail before it declares the remote gone and tosses up an error (I have not trolled the source..)--would it be possible to configure both the frequency and number of failures at some point?
I guess I should also ask: what additional info is helpful?  Running debug with the cluster active produces a lot of output, IIRC, the cluster stays connected, so if I'm not right there when it loses the sql db, I haven't captured the other messages.  I'll redir to a file and share that once I'm in the shack. 
Otherwise, no real issues at all; it works wonderfully and I enjoy using it.

KC0VKN
Additonal info

<p>I've been watching this more; I was in the process of QSY to a new QTH and using my original log over WAN. &nbsp;It appears to only happen when the cluster is used, I won't lose connection to my LAN connected database unless something outside of CQRlog happens; but if the cluster is active, eventually it will complain of lost connection.</p>
<p>I don't know if this happens to a locally (as in, on the same host as cqrlog is being run), but I would expect if it did, there would be other reports.</p>
<p>&nbsp;</p>
<p>Regardless; I'm fully converted to CQRlog. &nbsp;It is outstanding.</p>