retry the quick connect startup probe, and exit non-zero when a start fails
This commit is contained in:
+4
-3
@@ -11,9 +11,10 @@ using Xunit;
|
||||
namespace Jellyfin.Server.Tests.HighAvailability;
|
||||
|
||||
/// <summary>
|
||||
/// A Redis store that cannot be reached degrades silently: the client is configured not to abort the
|
||||
/// connection and every call site swallows failures. The probe is the only startup signal, so both of
|
||||
/// its outcomes are pinned here.
|
||||
/// A transcode store that cannot be reached degrades silently: the client is configured not to abort
|
||||
/// the connection and every call site swallows failures. Quick connect's startup read is what stops a
|
||||
/// pod coming up against a dead store; this probe only reports it, so both of its outcomes are pinned
|
||||
/// here - including that it never throws, which is what keeps the two policies from competing.
|
||||
/// </summary>
|
||||
public sealed class TranscodeStoreConnectivityProbeTests
|
||||
{
|
||||
|
||||
Reference in New Issue
Block a user