[prev in list] [next in list] [prev in thread] [next in thread] 

List:       intermezzo-devel
Subject:    Known problem ?
From:       Pascal Lengard <pascal.lengard () wanadoo ! fr>
Date:       2002-05-15 11:43:39
[Download RAW message or body]

Hi,

I'am still testing intermezzo.
using two boxes (node1 and node2), running redhat 7.2 with all available updates,
and a standard (kernel.org) linux kernel 2.4.18 with intersync 0.9.4-1 installed
via rpm.
node1 runs intersync like this:
  intersync /exports/data
node2 runs intersync like this:
  intersync --server=node1 /imports/data


When I create a file on node1 via this command:
  echo "foobar" > /exports/data/foobar

On node2, I got messages on /var/log/message (result of printk), saying:
  kernel: (journal.c:presto_truncate_lml,l. 1424 870): truncate_lml error -5

This message does not prevent node2 to synchronize correctly, file foobar
gets transfered.


I added some printk to intermezzo module to get better view of the problem,
it appears that the first time I create the file, these transferts are logged:
  clustertest-node2 kernel: PLUG : 492 : presto_fread rc=56 len=56
  clustertest-node2 last message repeated 2 times
  clustertest-node2 kernel: PLUG : 492 : presto_fread rc=32 len=32
  clustertest-node2 kernel: PLUG : 492 : presto_fread rc=4 len=4
  clustertest-node2 last message repeated 2 times
  clustertest-node2 kernel: PLUG : 492 : presto_fread rc=16 len=16

this one is the source of the message: read 9 bytes, asked for 32
  clustertest-node2 kernel: PLUG : 492 : presto_fread rc=9 len=32
  clustertest-node2 kernel: (journal.c:presto_truncate_lml,l. 1424 870): truncate_lml \
error -5

if I replay the 'echo ....' command to recreate the file, same appears except
we are able to read more and more bytes (increments by 1 each time)

  clustertest-node2 kernel: PLUG : 492 : presto_fread rc=10 len=32
  clustertest-node2 kernel: PLUG : 492 : presto_fread rc=11 len=32
  ...
  clustertest-node2 kernel: PLUG : 492 : presto_fread rc=31 len=32
all accompagned with this error message:
  clustertest-node2 kernel: (journal.c:presto_truncate_lml,l. 1424 868): truncate_lml \
error -5

and when we finally are able to read 32 bytes as we try since the beginning,
we got:
  clustertest-node2 kernel: PLUG : 492 : presto_fread rc=32 len=32
    (no message since it worked)
  clustertest-node2 kernel: PLUG : 492 : presto_fread rc=4 len=4
  clustertest-node2 kernel: PLUG : 492 : presto_fread rc=16 len=16
  clustertest-node2 kernel: (journal.c:presto_last_record,l. 1002 869): InterMezzo: \
Warning long record tail at 5744, rec tail_offset at 0 (size 1021455450)

and next time, "counter" is restarted to 0:
  clustertest-node2 kernel: PLUG : 492 : presto_fread rc=0 len=32
  clustertest-node2 kernel: (journal.c:presto_truncate_lml,l. 1424 869): truncate_lml \
error -5




Is this "known" behaviour ??
seems like we always try to get lml thingy with 32 bytes chunks ?

Pascal



_______________________________________________________________

Have big pipes? SourceForge.net is looking for download mirrors. We supply
the hardware. You get the recognition. Email Us: bandwidth@sourceforge.net
_______________________________________________
intermezzo-devel mailing list
intermezzo-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/intermezzo-devel


[prev in list] [next in list] [prev in thread] [next in thread]