could not connect error

[mpeters]

Am I correct in assuming that if my apache error logs show
something like this....

(2)No such file or directory: Could not connect to spread with
private_name ap28474. Error -6

That mod_log_spread will try and reconnect. We are seeing this
error occasionally pop up in the logs. Are we losing any data?

[Dave Weis]

Am I correct in assuming that if my apache error logs show something like
this....
(2)No such file or directory: Could not connect to spread with private_name
ap28474. Error -6
That mod_log_spread will try and reconnect. We are seeing this error
occasionally pop up in the logs. Are we losing any data?

We saw this also, I changed ours to put in a loop and print a counter of
how many times it tried to reconnect. It usually did it once, sometimes
twice, but straightened out afterwards.

As far as I know it didn't lose data, though.

[Jose R. Ilabaca]

Every time a new apache child goes up this message appears in the log.
When you start apache you have several lines like those, also when a new
child is created. It doesn't mean really a problem, I remember this
question in the group some time ago and it was an "inofensive" problem.
JRI

Post by Dave Weis

Am I correct in assuming that if my apache error logs show something like
this....
(2)No such file or directory: Could not connect to spread with private_name
ap28474. Error -6
That mod_log_spread will try and reconnect. We are seeing this error
occasionally pop up in the logs. Are we losing any data?

We saw this also, I changed ours to put in a loop and print a counter of
how many times it tried to reconnect. It usually did it once, sometimes
twice, but straightened out afterwards.
As far as I know it didn't lose data, though.

[mpeters]

Thank you very much. Something else that may or may not be
related is that we set up some new boxes with mod_log_spread and
after they went into production we started to see a *lot*
(meaning 1 or 2 times a minute) of segmentation faults of child
processes. We looked at the previous error logs just to make
sure, and we didn't see these kind of problems. The other
application code (mod_perl stuff) hasn't changed. We turned off
the mod_log_spread config directives to check to see if it was
something along the way and we still got the same seg faults.

using apache 1.3.31
with mod_log_spread 1.04

We had to try a few tricks to get apache to compile right (find
the spread libs) with mod_ssl, mod_perl. Not sure if maybe we did
something to make apache unstable or not. I don't know if anyone
else has seen this or not or has mod_log_spread working with the
rest of the software I mentioned above. Just thought I would ask
so that would could eliminate one link in this chain.

Thanks

Post by Jose R. Ilabaca
Every time a new apache child goes up this message appears in the log.
When you start apache you have several lines like those, also when a new
child is created. It doesn't mean really a problem, I remember this
question in the group some time ago and it was an "inofensive" problem.
JRI

Post by Dave Weis

Am I correct in assuming that if my apache error logs show something like
this....
(2)No such file or directory: Could not connect to spread with private_name
ap28474. Error -6
That mod_log_spread will try and reconnect. We are seeing this error
occasionally pop up in the logs. Are we losing any data?

We saw this also, I changed ours to put in a loop and print a counter of
how many times it tried to reconnect. It usually did it once, sometimes
twice, but straightened out afterwards.
As far as I know it didn't lose data, though./

[mpeters]

Thank you very much. Something else that may or may not be related is
that we set up some new boxes with mod_log_spread and after they went
into production we started to see a *lot* (meaning 1 or 2 times a
minute) of segmentation faults of child processes. We looked at the
previous error logs just to make sure, and we didn't see these kind of
problems. The other application code (mod_perl stuff) hasn't changed. We
turned off the mod_log_spread config directives to check to see if it
was something along the way and we still got the same seg faults.
using apache 1.3.31
with mod_log_spread 1.04
We had to try a few tricks to get apache to compile right (find the
spread libs) with mod_ssl, mod_perl. Not sure if maybe we did something
to make apache unstable or not. I don't know if anyone else has seen
this or not or has mod_log_spread working with the rest of the software
I mentioned above. Just thought I would ask so that would could
eliminate one link in this chain.

Sorry, to post this to the list as it turned out to be
irrelevant. We traced the problem to call to Apache::warn causing
the seg faults. We'll be working with that further, but just
wanted to let everyone know that it wasn't mod_log_spread.

Thanks

Thanks

Post by Jose R. Ilabaca
Every time a new apache child goes up this message appears in the log.
When you start apache you have several lines like those, also when a
new child is created. It doesn't mean really a problem, I remember
this question in the group some time ago and it was an "inofensive"
problem.
JRI

Post by Dave Weis

Am I correct in assuming that if my apache error logs show something
like this....
(2)No such file or directory: Could not connect to spread with
private_name ap28474. Error -6
That mod_log_spread will try and reconnect. We are seeing this error
occasionally pop up in the logs. Are we losing any data?

We saw this also, I changed ours to put in a loop and print a counter
of how many times it tried to reconnect. It usually did it once,
sometimes twice, but straightened out afterwards.
As far as I know it didn't lose data, though./