connection refused
waiting for connection
If transfers succeed but an error appears in the logs, the storage client may be starting a transfer before the SDCX Server is able to accept connections. In this case, connection refused
errors should only occur for a few seconds until the SDCX Server becomes available and the transfer begins.
If transfers fail, these errors can indicate that:
An endpoint requires a custom transfer IP address
Verify that the transfer IP addresses are able to connect to each other over the network.
The SDCX Server is not running on the destination endpoint
After logging into the server via a terminal, run the status
command to verify that the Jet Server is running: systemctl status jet-server
A firewall is blocking the required ports
To isolate the issue:
netstat
command to check that the required ports are open and in a LISTEN
state:netstat -tnl
2. Telnet to the port over the local loopback adapter to ensure that incoming TCP connections on ports 443, 49221 and 50221 are being accepted:
telnet <host> <port_number>
3. Use the dds_cnctst -udp
command to test access to UDP port 49221.
4. If a waiting for connection
error remains unresolved, check that ports using bi-directional connections have sockets that reach an ESTABLISHED
state:
netstat -tan | grep <port_number>
If an outbound firewall rule is restricting one or more of the required ports, any inbound connection attempts will be partially accepted by the server but left in a SYN_SENT
or SYN_RECV
state because the firewall is not allowing the bi-directional TCP handshake to complete.
stale NFS file handle
This error message can appear if a file changes during the 10-second waiting period between the beginning of a transfer and when files start moving to the destination.
Once modifications to the file stop, the transfer will resume.
file does not exist
This error message may appear if the file on the source endpoint has been deleted, renamed or moved.
Check the source directory to ensure that the file still exists.
direct stat returned error
This error message indicates that Jet is unable to identify the status of files on the source or destination endpoint. This issue may occur if the endpoints on a route cannot connect, if storage is not connected to an endpoint, or if network storage permissions do not allow access to file status information.
To isolate the issue:
There may be an issue where a server does not appear in the Jet management console when configuring a Jet endpoint after a successful SDCX Server installation.
To isolate the issue, check that the deviceID
and keys appear in these files:
/usr/Signiant_Media_Shuttle/bin/tomcat/webapps/ROOT/WEB-INF/config.properties
/usr/local/etc/signiant/signiant-communication-service-config.json5
/usr/local/etc/signiant/signiant-communication-service-keys.json5
If no deviceID
is associated with the configuration or service key files, run the dds_admin
utility in the bin
directory to check connectivity. Verify that the time offset is not more than a few seconds off.
In some cases, Jet transfer speed may be limited by the number of worker threads allocated. By default, there are 8 worker threads available to pass chunks of data at a time. Increasing the number of worker threads to 16 and the part length to 16Mi on the source and destination endpoints may improve transfer speed.
To change the number of worker threads and the part length:
/usr/local/etc/signiant/jet-server/config.json
C:\Program Files\Signiant\Jet Server\config.json
storageClientOverrides
section to increase worker threads to 16
and part length to 16Mi
for the source and the target.
"storageClientOverrides": {
"enabled": true,
"flightGateway": "",
"source": {
"partLength": "16Mi",
"partRetryLimit": 10,
"workerThreads": 16,
"maxPartLength": "128Mi"
},
"target": {
"partLength": "16Mi",
"partRetryLimit": 10,
"workerThreads": 16,
"maxPartLength": "128Mi"
}
}
Save the configuration file and restart Jet services:
Linux: In your terminal, run >./siginit restart
.
Windows: In the Services control panel, right-click on the services and choose Restart.
Transfer some new files to determine if the changes have increased the transfer speed.
If there is no gain in speed, use iPerf to determine the maximum route transfer speed.