On one of our remote hubs the Attach queue, and consequently the correspondent GET queue that exists in the primary stopped working (showing disconnected and messages are piling up).
The hub.log (level 3) locally at the remote hub is showing the following error:
Jun 26 22:19:16:326 [5300] 0 hub: sxxxxxxxsf: read head failed, 103957.sds
Jun 26 22:19:51:891 [5524] 0 hub: Disconnecting queue route for 'Attach_QOS' for repeated or terminal failure (2,1)
Jun 26 22:21:49:969 [5524] 0 hub: Disconnecting queue route for 'Attach_QOS' for repeated or terminal failure (2,1)
[..]
Jun 27 23:07:49:847 [5528] 0 hub: Failed to send response to post request
Jun 27 23:07:49:925 [5524] 0 hub: Disconnecting queue route for 'Attach_QOS' for repeated or terminal failure (2,1)
[..]
Jun 29 08:34:56:853 [6192] 0 hub: Writing to SDS file (save) - queue file operation failed for 'Attach_Alarm' file 2657.sds; pos=3715012 (simultaneous), npds=3131
Jun 29 08:35:13:023 [6196] 0 hub: savepds: who=4 failed, error=28
Jun 29 08:35:13:023 [6196] 0 hub: Writing to SDS file (save) - queue file operation failed for 'Attach_QOS' file 104603.sds; pos=69803733 (non simultaneous), npds=112056
[..]
Jun 29 09:55:28:617 [7068] 0 hub: CORE exit
Jun 29 09:55:28:896 [6172] 0 hub: Gave up waiting to shut down post_thread 1 at address 000000000000048C
Jun 29 09:55:28:896 [6172] 0 hub: Gave up waiting to shut down post_thread 2 at address 0000000000000494
Jun 29 09:55:28:897 [6172] 0 hub: Gave up waiting to shut down post_thread 3 at address 0000000000000498
Jun 29 09:55:28:897 [6172] 0 hub: Gave up waiting to shut down post_thread 4 at address 000000000000049C
Jun 29 09:55:31:109 [6812] 0 hub: EXIT HUB: hub shutdown
Jun 29 09:55:31:109 [6812] 0 hub: Waiting for tunnel to terminate ...
Jun 29 09:55:31:109 [6812] 0 hub: Waited for 0 seconds for tunnel to terminate (OK)
Release : UIM 20.x, 9.x
Component : UIM - HUB
Possible corrupt SDS file
These errors may indicate that there is a corrupt SDS file that caused this issue.
In this case, the only option would be to delete the queue, restart the hub, and then re-make the queue.
After recreating the queue the data will be processed successfully.