You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Imagine the server should crash or be killed and rebooted. In this case the clients are going to restore their acquired locks. However they will do so only after their heartbeat receives the Unknown Peer error from the server. Should to many clients acquire new locks, before this happens, the whole system runs at risk of increasing the semaphore count beyond max. While this is intentionally supported, it would be nice if we could delay handing out new locks after a server reboot. Specifically increasing it two a time somewhat longer than the heartbeat interval should make incrementing the semaphore count beyond max nearly impossible.
The text was updated successfully, but these errors were encountered:
Imagine the server should crash or be killed and rebooted. In this case the clients are going to restore their acquired locks. However they will do so only after their heartbeat receives the
Unknown Peer
error from the server. Should to many clients acquire new locks, before this happens, the whole system runs at risk of increasing the semaphore count beyond max. While this is intentionally supported, it would be nice if we could delay handing out new locks after a server reboot. Specifically increasing it two a time somewhat longer than the heartbeat interval should make incrementing the semaphore count beyond max nearly impossible.The text was updated successfully, but these errors were encountered: