I received this in my log:
StreamState[Open]; DataState[Suspect]; StatusCode[None]; StatusText[Comms Outage]; Symbol[ES2430C7]
How should I proceed? Does it mean that I should failover to my secondary connection?
The above log entry means that the quality of the data is suspect, the stream is still open, but something has gone wrong upstream and therefore, the server is advising you not to rely on the data for ES2430C7 - it may not be fully up to date / accurate.
When the comms recovers you will receive a Refresh (unsolicited) containing the latest data and DataState[OK].
Please expand on what you mean by secondary connection. What is your primary and secondary connection? Are you connecting to ADS or using EMS?
Also, how do you plan to switch to secondary? It may not always help. e.g. if you are connecting to an ADS and you switch to a secondary ADS it is quite possible that you would not get non-Suspect data on the secondary if the upstream servers for both are experiencing comms issues.
Just on the related matter of secondary server and failover etc - which API are you using? If you are using EMA then you should use ChannelSets if you are not already doing so. This will allow the API itself to attempt failover to a secondary server if connection to the primary is lost. Just to be clear, the log message above does not indicate a lost connection for your application - the StreamState[Open] confirms that your a connection to your server is still open and event stream for the individual instrument is still open.
We are writing RFA to direct connect the EDGE devices. Our secondary is an independent connection to a different EDGE device that connects via a different circuit to a different data center. So totally redundant. I just want to confirm that if my feedhandler receive such messages (StreamState[Open]; DataState[Suspect]; StatusText[Comms Outage]), what should be the appropriate action. In the case, I think I should switch to the other connection that is good.
Well, if the outage is between the active EDGE and the Distribution POP, then switching to the secondary EDGE should help - assuming the comms for the other EDGE is fine.
However, if the comms outage was higher up the chain e.g. between Reuters and an exchange, then switching would not help.
Were there any other status messages e.g. at the Service level? I would expect that if the outage was between the EDGE and Dist Pop then the Elektron Service(s) would be marked as down and Suspect DataState for all open streams.
If, however, the suspect DataState was only communicated for only part of your watchlist, then I would expect the outage was limited to the source of those instruments e.g. a particular exchange etc.