When logged into BADBOX I type
"rsync -e ssh -r SMALL-DIR [EMAIL PROTECTED]:/tmp"
and it works fine as long as the SMALL-DIR directory has less than 2 dozen files and the files are less than 32KB in size.
As soon as I do "rsync -e ssh -r BIG-DIR [EMAIL PROTECTED]:/tmp"
I get the following output:
erroring writing 32768 bytes - exiting
On the "otherbox" in /tmp there is a BIG-DIR, and almost all of the files are there except for the BIG files.
I ran "strace rsync ..." and it looks like everything goes fine on all the small files, then it gets to the big file "TEST/linux.tar.gz" and
see here:
...
write(1, "sender finished TEST/licence.o\n", 31sender finished TEST/licence.o
) = 31
select(6, [5], NULL, NULL, {60, 0}) = 1 (in [5], left {60, 0})
read(5, "%\0\0\t", 4) = 4
select(6, [5], NULL, NULL, {60, 0}) = 1 (in [5], left {60, 0})
read(5, "recv_generator(TEST/linux.tar.gz"..., 37) = 37
write(1, "recv_generator(TEST/linux.tar.gz"..., 37recv_generator(TEST/linux.tar.gz,19)
) = 37
select(6, [5], NULL, NULL, {60, 0}) = 1 (in [5], left {60, 0})
read(5, "\20\0\0\7", 4) = 4
select(6, [5], NULL, NULL, {60, 0}) = 1 (in [5], left {60, 0})
read(5, "\23\0\0\0", 4) = 4
write(1, "send_files(19,TEST/linux.tar.gz)"..., 33send_files(19,TEST/linux.tar.gz)
) = 33
select(6, [5], NULL, NULL, {60, 0}) = 1 (in [5], left {60, 0})
read(5, "\0\0\0\0", 4) = 4
select(6, [5], NULL, NULL, {60, 0}) = 1 (in [5], left {60, 0})
read(5, "\274\2\0\0", 4) = 4
select(6, [5], NULL, NULL, {60, 0}) = 1 (in [5], left {60, 0})
read(5, "\0\0\0\0", 4) = 4
open("TEST/linux.tar.gz", O_RDONLY|O_LARGEFILE) = 3
fstat64(3, {st_mode=S_IFREG|0640, st_size=1602359, ...}) = 0
write(1, "send_files mapped TEST/linux.tar"..., 52send_files mapped TEST/linux.tar.gz of size 1602359
) = 52
select(5, NULL, [4], NULL, {60, 0}) = 1 (out [4], left {60, 0})
write(4, "\23\0\0\0", 4) = 4
select(5, NULL, [4], NULL, {60, 0}) = 1 (out [4], left {60, 0})
write(4, "\0\0\0\0", 4) = 4
select(5, NULL, [4], NULL, {60, 0}) = 1 (out [4], left {60, 0})
write(4, "\274\2\0\0", 4) = 4
select(5, NULL, [4], NULL, {60, 0}) = 1 (out [4], left {60, 0})
write(4, "\0\0\0\0", 4) = 4
write(1, "calling match_sums TEST/linux.ta"..., 37calling match_sums TEST/linux.tar.gz
) = 37
write(1, "TEST/linux.tar.gz\n", 18TEST/linux.tar.gz
) = 18
select(5, NULL, [4], NULL, {60, 0}) = 1 (out [4], left {60, 0})
write(4, "\0\200\0\0", 4) = 4
mmap2(NULL, 266240, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0x401a0000
read(3, "\37\213\10\10\317`\177=\0\3bp_inst_linux.tar\0\354=Yp"..., 262144) = 262144
select(5, NULL, [4], NULL, {60, 0}) = 1 (out [4], left {60, 0})
write(4, "\37\213\10\10\317`\177=\0\3bp_inst_linux.tar\0\354=Yp"..., 32768) = -1 ENOBUFS (No buffer space available)
write(2, "erroring writing 32768 bytes - e"..., 39erroring writing 32768 bytes - exiting
) = 39
rt_sigaction(SIGUSR1, {SIG_IGN}, {0x8050420, [USR1], SA_RESTART|0x4000000}, 8) = 0
rt_sigaction(SIGUSR2, {SIG_IGN}, {0x8050440, [USR2], SA_RESTART|0x4000000}, 8) = 0
getpid() = 20601
kill(20602, SIGUSR1) = 0
--- SIGCHLD (Child exited) ---
munmap(0x40018000, 4096) = 0
_exit(12) = ?
...
I read about "ENOBUFS" which led me to try adjusting the TCP socket buffer sizes.... I tried the following (this box has 2GB of RAM)
/proc/sys/fs/files-nr: 3050 1406 40960 /proc/sys/net/core/rmem_max: 128388607 /proc/sys/net/core/wmem_max: 128388607 /proc/sys/net/ipv4/tcp_adv_win_scale: 2 /proc/sys/net/ipv4/tcp_window_scaling: 1 /proc/sys/net/ipv4/tcp_rmem: 8192 128000 128388607 /proc/sys/net/ipv4/tcp_wmem: 8192 112000 127388607
This box does have 7250 open file descriptors reported by "lsof" But the box isn't that busy, it has a load of only about 0.5
cat /proc/meminfo total: used: free: shared: buffers: cached: Mem: 2108256256 1899868160 208388096 2043904 190640128 1422811136 Swap: 6440263680 185999360 6254264320 MemTotal: 2058844 kB MemFree: 203504 kB MemShared: 1996 kB Buffers: 186172 kB Cached: 1298792 kB SwapCached: 90672 kB Active: 1071248 kB Inact_dirty: 214688 kB Inact_clean: 291696 kB Inact_target: 514692 kB HighTotal: 1179584 kB HighFree: 2036 kB LowTotal: 879260 kB LowFree: 201468 kB SwapTotal: 6289320 kB SwapFree: 6107680 kB BigPagesFree: 0 kB
vmstat 2 3
procs memory swap io system cpu
r b w swpd free buff cache si so bi bo in cs us sy id
2 0 0 181640 203504 186200 1299104 0 0 1 1 0 0 0 1 0
0 0 0 181640 203504 186200 1299112 0 0 0 0 142 228 24 0 75
2 0 0 181640 203504 186204 1299120 0 0 0 54 184 342 2 0 98
Does anybody have any ideas?
Thanks in advance, -Ben.
-- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
