mirror of
https://github.com/wassname/ray.git
synced 2026-07-10 17:07:56 +08:00
Check /dev/shm on Linux (#174)
* check available shared memory when starting object store * exit with error if not enough shared memory available for object store * Some comments and formatting.
This commit is contained in:
committed by
Robert Nishihara
parent
431bba3c8a
commit
b1e76e582e
@@ -254,8 +254,20 @@ def start_objstore(node_ip_address, redis_address, cleanup=True, redirect_output
|
||||
# memory. To not overflow it, we set the plasma memory limit to 0.4 times
|
||||
# the size of the physical memory.
|
||||
objstore_memory = int(system_memory * 0.4)
|
||||
# Compare the requested memory size to the memory available in /dev/shm.
|
||||
shm_fd = os.open("/dev/shm", os.O_RDONLY)
|
||||
try:
|
||||
shm_fs_stats = os.fstatvfs(shm_fd)
|
||||
# The value shm_fs_stats.f_bsize is the block size and the value
|
||||
# shm_fs_stats.f_bavail is the number of available blocks.
|
||||
shm_avail = shm_fs_stats.f_bsize * shm_fs_stats.f_bavail
|
||||
if objstore_memory > shm_avail:
|
||||
print("Warning: Reducing object store memory because /dev/shm has only {} bytes available. You may be able to free up space by deleting files in /dev/shm. If you are inside a Docker container, you may need to pass an argument with the flag '--shm-size' to 'docker run'.".format(shm_avail))
|
||||
objstore_memory = int(shm_avail * 0.8)
|
||||
finally:
|
||||
os.close(shm_fd)
|
||||
else:
|
||||
objstore_memory = int(system_memory * 0.75)
|
||||
objstore_memory = int(system_memory * 0.8)
|
||||
# Start the Plasma store.
|
||||
plasma_store_name, p1 = plasma.start_plasma_store(plasma_store_memory=objstore_memory, use_profiler=RUN_PLASMA_STORE_PROFILER, redirect_output=redirect_output)
|
||||
# Start the plasma manager.
|
||||
|
||||
@@ -13,8 +13,10 @@
|
||||
#include <stdio.h>
|
||||
#include <stdlib.h>
|
||||
#include <unistd.h>
|
||||
#include <fcntl.h>
|
||||
#include <sys/ioctl.h>
|
||||
#include <sys/socket.h>
|
||||
#include <sys/statvfs.h>
|
||||
#include <sys/types.h>
|
||||
#include <sys/un.h>
|
||||
#include <getopt.h>
|
||||
@@ -664,6 +666,26 @@ int main(int argc, char *argv[]) {
|
||||
if (system_memory == -1) {
|
||||
LOG_FATAL("please specify the amount of system memory with -m switch");
|
||||
}
|
||||
#ifdef __linux__
|
||||
/* On Linux, check that the amount of memory available in /dev/shm is large
|
||||
* enough to accommodate the request. If it isn't, then fail. */
|
||||
int shm_fd = open("/dev/shm", O_RDONLY);
|
||||
struct statvfs shm_vfs_stats;
|
||||
fstatvfs(shm_fd, &shm_vfs_stats);
|
||||
/* The value shm_vfs_stats.f_bsize is the block size, and the value
|
||||
* shm_vfs_stats.f_bavail is the number of available blocks. */
|
||||
int64_t shm_mem_avail = shm_vfs_stats.f_bsize * shm_vfs_stats.f_bavail;
|
||||
close(shm_fd);
|
||||
if (system_memory > shm_mem_avail) {
|
||||
LOG_FATAL(
|
||||
"System memory request exceeds memory available in /dev/shm. The "
|
||||
"request is for %" PRId64 " bytes, and the amount available is %" PRId64
|
||||
" bytes. You may be able to free up space by deleting files in "
|
||||
"/dev/shm. If you are inside a Docker container, you may need to pass "
|
||||
"an argument with the flag '--shm-size' to 'docker run'.",
|
||||
system_memory, shm_mem_avail);
|
||||
}
|
||||
#endif
|
||||
LOG_DEBUG("starting server listening on %s", socket_name);
|
||||
start_server(socket_name, system_memory);
|
||||
}
|
||||
|
||||
Reference in New Issue
Block a user