Does anyone have any experience troubleshooting an error 174 (media manager -
system error)? The error that is in the problems log is: "attempt to write 0
gigabytes, 166527160 bytes, it is not a multiple of 512 bytes". This only
happens with one client.
I patched the master/media server and all clients from 4.5MP3 to 4.5MP5, no
resolution. The master/media server is Solaris, the client in question is
Linux. The backup data stream is going over the 100Mb network. I turned on
bpcd logging on the client and bptm logging on the server. Could not find
any error output in bpcd on the client. On the server, these messages show
up:
20:07:52.857 [2134] <2> write_data: received first buffer (32768 bytes), begin
writing data
20:08:32.654 [2146] <2> fill_buffer: [2134] socket is closed, waited for empty
buffer 263 times, delayed 791 times, read 166527160 bytes
20:08:32.663 [2134] <2> write_data: writing block shorter than BUFF_SIZE, 184
bytes
20:08:32.664 [2134] <2> write_data: writing short block, 184 bytes, remainder
184
20:08:32.664 [2134] <2> wait_for_sigcld: waiting for child to exit, timeout is
300
20:08:37.664 [2134] <2> child_wait: SIGCLD: exit=174, signo=0 core=no,
pid=2146
20:08:37.664 [2134] <2> write_data: waited for full buffer 189 times, delayed
295 times
20:08:37.664 [2134] <2> write_backup: write_data() returned, exit_status =
174, CINDEX = 0, TWIN_INDEX = 0, backup_status = -8
20:08:37.665 [2134] <2> write_backup: tp.tv_sec = 1069290512, stp.tv_sec =
1069290472, tp.tv_usec = 664864, stp.tv_usec = 857639, et = 39808,
mpx_total_kbytes[TWIN_INDEX = 0] = 162624
Has anyone ever dealt with this issue before? This client I believe was
working properly before, but I cannot find anything that might have changed.
Is it something that could be fixed by altering buffer sizes, or is it a
network issue?
I appreciate any assistance rendered.
--
Doug Bell
Consultant / SysAdmin
Perfect Order
dbell AT perfectorder DOT com
|