Liking cljdoc? Tell your friends :D

alda.server


*no-system-exit*clj

source

add-or-requeue-workerclj

(add-or-requeue-worker address)
source

all-workersclj

(all-workers)
source

all-workers-are-busy-responseclj

source

available-workersclj

source

blacklist-worker!clj

(blacklist-worker! address)
source

busy-workersclj

source

cycle-workers!clj

(cycle-workers! backend port workers)
source

exit!clj

(exit! exit-code)
source

fire-lazy-workers!clj

(fire-lazy-workers!)
source

HEARTBEAT-INTERVALclj

source

lay-off-worker!clj

(lay-off-worker!)

Lays off the last worker in the queue. Bad luck on its part.

To 'lay off' a worker, we cannot simply send a KILL message, otherwise the worker may die without finishing its workload (e.g. if it's in the middle of playing a score). We want the worker to finish what it's doing and then shut down.

To do this, we...

  • remove it from the queue
  • temporarily 'blacklist' its address so it won't be re-queued the next time it sends a heartbeat
  • in a separate thread, wait a safe amount of time and then remove the address from the blacklist; this is to keep the blacklist from growing too big over time

Once the worker is out of the queue and we aren't letting it back in, the worker will stop getting heartbeats from the server and shut itself down.

Lays off the last worker in the queue. Bad luck on its part.

To 'lay off' a worker, we cannot simply send a KILL message, otherwise the
worker may die without finishing its workload (e.g. if it's in the middle of
playing a score). We want the worker to finish what it's doing and then shut
down.

To do this, we...
- remove it from the queue
- temporarily 'blacklist' its address so it won't be re-queued the next time
  it sends a heartbeat
- in a separate thread, wait a safe amount of time and then remove the
  address from the blacklist; this is to keep the blacklist from growing too
  big over time

Once the worker is out of the queue and we aren't letting it back in, the
worker will stop getting heartbeats from the server and shut itself down.
sourceraw docstring

murder-workers!clj

(murder-workers! backend)
source

no-workers-available-responseclj

source

note-that-worker-is-busyclj

(note-that-worker-is-busy address)
source

pong-responseclj

source

remove-worker-from-queueclj

(remove-worker-from-queue address)
source

running?clj

source

shut-down!clj

(shut-down! backend)
source

shutting-down-responseclj

source

start-server!clj

(start-server! workers frontend-port & [verbose?])
source

start-workers!clj

(start-workers! workers port)
source

status-responseclj

(status-response available total backend-port)
source

successful-responseclj

source

supervise-workers!clj

(supervise-workers! port desired)

Ensures that there are at least desired number of workers available by counting how many we have and starting more if needed.

Ensures that there are at least `desired` number of workers available by
counting how many we have and starting more if needed.
sourceraw docstring

SUSPENDED-INTERVALclj

source

unsuccessful-responseclj

source

version-responseclj

source

workerclj

(worker address)
source

worker-blacklistclj

source

WORKER-CHECK-INTERVALclj

source

worker-expiration-dateclj

(worker-expiration-date)
source

WORKER-LIVESclj

source

WORKER-START-INTERVALclj

source

cljdoc is a website building & hosting documentation for Clojure/Script libraries

× close