Adam Lugowski

I'm experiencing the same problem. I think it has to do with having multiple checks. `pass()` constructs the check ID to be `"service:" + serviceId`, but if you have multiple...

I found a workaround. `RegCheck`, the class used to bundle check registration with service registration, doesn't allow you to specify IDs, and for whatever reason sometimes it auto selects an...

@rickfast and Orbitz team: I think it makes sense for the `check` specified in registration to have a forced id of `"service:" + serviceId` to make sure `pass()` works properly....

@rickfast I see, no worries. This fix would only be for future users. Registering checks separately works better for my use case anyway, so I'm not blocked. I'm not sure...

@SuppieRK that is a separate issue that warrants its own ticket. This ticket is about a name mismatch only.

This already exists. Consul keeps track of this timer for health checks, not service instances. You can add a deregisterCriticalServiceAfter to any health check: Registration.RegCheck deadManSwitch = ImmutableRegCheck.builder() .ttl("10s") .deregisterCriticalServiceAfter("3h")...

You're right, I also didn't realize how much memory `inplace_merge` might actually allocate. Is that too much for your application? `pluggable_sort` also lets you replace inplace_merge with another implementation. I...

I went down an `inplace_merge` rabbit hole, and found something that seems usable. [This SO post](https://stackoverflow.com/a/4375732) references [an implementation](http://keithschwarz.com/interesting/code/?dir=inplace-merge) of the "Practical In-Place Merging" paper. That code only supports merging...

Good news! I added a [quicksort](https://github.com/alugowski/poolSTL/pull/29) implementation that appears to work a lot better than the mergesort we've been trying. Unlike merging, partitioning is both fast and does not need...

Great! I'm glad it's working well for you! > There seems to be some small overhead now when calling `pluggable_sort` and using a single thread. Good point, that is an...