imdiff: No space left on device
104 views
Skip to first unread message
Chat Hull
Dec 20, 2013, 1:19:07 PM12/20/13
to Astro MIRIAD Users
[Below is a message from Statia, who's having trouble posting. Please reply-all.]
Hi MIRIAD users,I’m trying to use imdiff to compare two maps, and when I do I get the following error:Imdiff: version 09-Feb-09Varied: amplitude,offset,x-shift,y-shift,expansion### Warning [imdiff]: Error writing to scratch file; check your $TMPDIR### Fatal Error [imdiff]: No space left on device
Also a window pops up that says “Low Disk Space Terminal has detected that the system disk is running low on swap space. Scrollback buffers may automatically be trimmed to conserved memory.”I’m at a loss because I have over 300 GB free on my hard drive. I can make all sorts of large files in other ways without a problem. I checked to make sure I can write stuff to $TMPDIR (I set $TMPDIR to /tmp and copied some large files to it). I realize that there’s a good chance this isn’t a miriad problem but I’m out of ideas, and my googling skills haven’t turned anything up. Suggestions appreciated!Best,Statia
---------- Forwarded message ----------
From: Statia H Cook <shc...@amnh.org>
Date: Fri, Dec 20, 2013 at 8:42 AM
Subject: Fwd: imdiff: No space left on device
To: Chat Hull <ch...@astro.berkeley.edu>
From: Statia H Cook <shc...@amnh.org>
Date: Fri, Dec 20, 2013 at 8:42 AM
Subject: Fwd: imdiff: No space left on device
To: Chat Hull <ch...@astro.berkeley.edu>
Hey Chat,
I was trying to email the Miriad google group with a problem but I can't send from my berkeley email address and its taking forever to get another email address added to the google group. I thought it might be faster if you posted the message for me- do
you mind?
The message is below.
Thanks! Statia
Begin forwarded message:
Begin forwarded message:
From: Statia Cook <shc...@amnh.org>
Date: December 19, 2013, 4:51:37 PM EST
To: <mir...@googlegroups.com>
Subject: imdiff: No space left on device
Hi MIRIAD users,
I’m trying to use imdiff to compare two maps, and when I do I get the following error:
Imdiff: version 09-Feb-09Varied: amplitude,offset,x-shift,y-shift,expansion### Warning [imdiff]: Error writing to scratch file; check your $TMPDIR### Fatal Error [imdiff]: No space left on device
Also a window pops up that says “Low Disk Space Terminal has detected that the system disk is running low on swap space. Scrollback buffers may automatically be trimmed to conserved memory.”
I’m at a loss because I have over 300 GB free on my hard drive. I can make all sorts of large files in other ways without a problem. I checked to make sure I can write stuff to $TMPDIR (I set $TMPDIR to /tmp and copied some large files to it). I realize that there’s a good chance this isn’t a miriad problem but I’m out of ideas, and my googling skills haven’t turned anything up. Suggestions appreciated!
Best,
Statia
melvyn wright
Dec 20, 2013, 3:51:53 PM12/20/13
to mir...@googlegroups.com, Chat Hull, Statia H Cook, Peter Teuben
Hi Statia, etal,
The message is coming from
melvyn-wrights-macbook-pro-4:~ wright$ grep -i TMPDIR $MIRSUBS/*.c
/Users/wright/miriad/src/subs/dio.c:/* 26-Oct-95 rjs Honour TMPDIR environment variable, if set. */
/Users/wright/miriad/src/subs/dio.c: "scratch" files are using $TMPDIR, if present, else current.
/Users/wright/miriad/src/subs/dio.c: s = getenv("TMPDIR");
/Users/wright/miriad/src/subs/hio.c: 'scratch' files are using $TMPDIR, if present, else current.
/Users/wright/miriad/src/subs/scrio.c: the $TMPDIR environment variable points to another directory.
/Users/wright/miriad/src/subs/scrio.c: bug_c( 'w',"Error opening scratch file; check your $TMPDIR");
/Users/wright/miriad/src/subs/scrio.c: bug_c( 'w',"Error closing scratch file; check your $TMPDIR");
/Users/wright/miriad/src/subs/scrio.c: bug_c( 'w',"Error reading from scratch file; check your $TMPDIR");
/Users/wright/miriad/src/subs/scrio.c: bug_c( 'w',"Error writing to scratch file; check your $TMPDIR");
You could try imdiff with smaller images and see if that works, but I don't
think that is the problem.
Peter Teuben might have some ideas.
cheers,
Mel.
--
You received this message because you are subscribed to the Google Groups "miriad" group.
To unsubscribe from this group and stop receiving emails from it, send an email to miriad+un...@googlegroups.com.
To post to this group, send email to mir...@googlegroups.com.
Visit this group at http://groups.google.com/group/miriad.
For more options, visit https://groups.google.com/groups/opt_out.
Peter Williams
Dec 20, 2013, 5:20:43 PM12/20/13
to mir...@googlegroups.com, Chat Hull, Statia H Cook, Peter Teuben
Hi,
I got in touch with Statia separately and got a dtruss log from her.
This appears to be another Mac 64-bit problem. Here's an (edited)
sequence of system calls:
open("scratch1.6141\0", 0x602, 0x1A4) = 6 0
lseek(0x6, 0x0, 0x2) = 0 0
unlink("scratch1.6141\0", 0x0, 0x0) = 0 0
lseek(0x6, 0x0, 0x0) = 0 0
write(0x6, "\0", 0x4B4) = 1204 0
lseek(0x6, 0x4B4000004B4, 0x0) = 1204 0
... ^^^^^^^^^^^^^
write(0x6, "\0", 0x4B4) = -1 Err#28 [= ENOSPC]
Basically MIRIAD tries to seek to a ridiculously large offset, then
write, which causes the no-space error.
Statia, it looks like you're using the MacPorts MIRIAD install. You
might try the CARMA binary distribution -- there have been times when
the different builds have had different bugs because some of these
things depend on how you compile the code.
Peter
I got in touch with Statia separately and got a dtruss log from her.
This appears to be another Mac 64-bit problem. Here's an (edited)
sequence of system calls:
open("scratch1.6141\0", 0x602, 0x1A4) = 6 0
lseek(0x6, 0x0, 0x2) = 0 0
unlink("scratch1.6141\0", 0x0, 0x0) = 0 0
lseek(0x6, 0x0, 0x0) = 0 0
write(0x6, "\0", 0x4B4) = 1204 0
lseek(0x6, 0x4B4000004B4, 0x0) = 1204 0
... ^^^^^^^^^^^^^
write(0x6, "\0", 0x4B4) = -1 Err#28 [= ENOSPC]
Basically MIRIAD tries to seek to a ridiculously large offset, then
write, which causes the no-space error.
Statia, it looks like you're using the MacPorts MIRIAD install. You
might try the CARMA binary distribution -- there have been times when
the different builds have had different bugs because some of these
things depend on how you compile the code.
Peter
Peter Teuben
Dec 20, 2013, 5:26:51 PM12/20/13
to mir...@googlegroups.com, Statia H Cook
Chat/Statia,
���� I'm on travel in Holland and due a very Murphyish chain of events had not been able to use my laptop for several days, and now looking at 250+ emails.� One of them was Statia's attempt to become
a miriad@googlegroups member, which somehow I could not make work on an android phone. This will be correctly shortly now that I have all the needed cables, a wifi booster and a charged battery.
Once you're a member, you will see from old emails that this is a "frequent" problem, and has to do with one of those loose ends of the 32-to-64 conversion process that we thought we had all covered by the school13 release.�� I just confirmed this problem also occurs for my on my Ubuntu13.02 laptop , so that should make a fix coming up even sooner.
peter
On 12/20/2013 01:19 PM, Chat Hull wrote:
���� I'm on travel in Holland and due a very Murphyish chain of events had not been able to use my laptop for several days, and now looking at 250+ emails.� One of them was Statia's attempt to become
a miriad@googlegroups member, which somehow I could not make work on an android phone. This will be correctly shortly now that I have all the needed cables, a wifi booster and a charged battery.
Once you're a member, you will see from old emails that this is a "frequent" problem, and has to do with one of those loose ends of the 32-to-64 conversion process that we thought we had all covered by the school13 release.�� I just confirmed this problem also occurs for my on my Ubuntu13.02 laptop , so that should make a fix coming up even sooner.
peter
On 12/20/2013 01:19 PM, Chat Hull wrote:
[Below is a message from Statia, who's having trouble posting. �Please reply-all.]
Hi MIRIAD users,�
I�m trying to use imdiff to compare two maps, and when I do I get the following error:
Imdiff: version 09-Feb-09Varied:� amplitude,offset,x-shift,y-shift,expansion### Warning [imdiff]:� Error writing to scratch file; check your $TMPDIR### Fatal Error [imdiff]:� No space left on device
Also a window pops up that says �Low Disk Space �Terminal has detected that the system disk is running low on swap space. Scrollback buffers may automatically be trimmed to conserved memory.��
I�m at a loss because I have over 300 GB free on my hard drive. I can make all sorts of large files in other ways without a problem. I checked to make sure I can write stuff to $TMPDIR (I set $TMPDIR to /tmp and copied some large files to it). I realize that there�s a good chance this isn�t a miriad problem but I�m out of ideas, and my googling skills haven�t turned anything up. Suggestions appreciated!
Best,�
Statia
---------- Forwarded message ----------
From: Statia H Cook <shc...@amnh.org>
Date: Fri, Dec 20, 2013 at 8:42 AM
Subject: Fwd: imdiff: No space left on device
To: Chat Hull <ch...@astro.berkeley.edu>
Hey Chat,�I was trying to email the Miriad google group with a problem but I can't send from my berkeley email address and its taking forever to get another email address added to the google group. I thought it might be faster if you posted the message for me- do you mind?�
The message is below.�
Thanks! Statia
Begin forwarded message:
From: Statia Cook <shc...@amnh.org>
Date: December 19, 2013, 4:51:37 PM EST
To: <mir...@googlegroups.com>
Subject: imdiff: No space left on device
Hi MIRIAD users,�
I�m trying to use imdiff to compare two maps, and when I do I get the following error:
Varied:� amplitude,offset,x-shift,y-shift,expansion### Warning [imdiff]:� Error writing to scratch file; check your $TMPDIR### Fatal Error [imdiff]:� No space left on device
Also a window pops up that says �Low Disk Space �Terminal has detected that the system disk is running low on swap space. Scrollback buffers may automatically be trimmed to conserved memory.��
I�m at a loss because I have over 300 GB free on my hard drive. I can make all sorts of large files in other ways without a problem. I checked to make sure I can write stuff to $TMPDIR (I set $TMPDIR to /tmp and copied some large files to it). I realize that there�s a good chance this isn�t a miriad problem but I�m out of ideas, and my googling skills haven�t turned anything up. Suggestions appreciated!
Best,�
Statia
--
Chat Hull
Graduate Student
Radio Astronomy Laboratory
UC Berkeley
LinkedIn:�linkedin.com/in/chathullWeb:�astro.berkeley.edu/~chat
Peter Teuben
Dec 20, 2013, 5:58:07 PM12/20/13
to mir...@googlegroups.com
peter,
i think that's about right, but i also have another patch from the
ATNF due,
which i'll apply. I'll test it and commit in the next hour.
other peter
i think that's about right, but i also have another patch from the
ATNF due,
which i'll apply. I'll test it and commit in the next hour.
other peter
melvyn wright
Dec 20, 2013, 5:59:05 PM12/20/13
to mir...@googlegroups.com, Chat Hull, Statia H Cook, Peter Teuben
Hi Peter,
Thanks for diagnosing and fixing this !
Best wishes for 2014,
Mel.
Peter Teuben
Dec 20, 2013, 6:58:04 PM12/20/13
to mir...@googlegroups.com
i've retrofitted imdiff from a series
of atnf updates, which was captured by Peter W in
his email as well, except for sanity i'll follow the ATNF reformatting rules.
I've confirmed it works.
For old style, do this:
��� mirboss
��� cd $MIRPROG/analysis
��� cvs update
��� mir.prog imdiff
and you should be good to go.
- peterT
his email as well, except for sanity i'll follow the ATNF reformatting rules.
I've confirmed it works.
For old style, do this:
��� mirboss
��� cd $MIRPROG/analysis
��� cvs update
��� mir.prog imdiff
and you should be good to go.
- peterT
On 12/20/2013 01:19 PM, Chat Hull wrote:
[Below is a message from Statia, who's having trouble posting. �Please reply-all.]
Varied:� amplitude,offset,x-shift,y-shift,expansion### Warning [imdiff]:� Error writing to scratch file; check your $TMPDIR### Fatal Error [imdiff]:� No space left on device
Also a window pops up that says �Low Disk Space �Terminal has detected that the system disk is running low on swap space. Scrollback buffers may automatically be trimmed to conserved memory.��
I�m at a loss because I have over 300 GB free on my hard drive. I can make all sorts of large files in other ways without a problem. I checked to make sure I can write stuff to $TMPDIR (I set $TMPDIR to /tmp and copied some large files to it). I realize that there�s a good chance this isn�t a miriad problem but I�m out of ideas, and my googling skills haven�t turned anything up. Suggestions appreciated!
Best,�
Statia
---------- Forwarded message ----------
From: Statia H Cook <shc...@amnh.org>
Date: Fri, Dec 20, 2013 at 8:42 AM
Subject: Fwd: imdiff: No space left on device
To: Chat Hull <ch...@astro.berkeley.edu>
Hey Chat,�I was trying to email the Miriad google group with a problem but I can't send from my berkeley email address and its taking forever to get another email address added to the google group. I thought it might be faster if you posted the message for me- do you mind?�
The message is below.�
Thanks! Statia
Begin forwarded message:
From: Statia Cook <shc...@amnh.org>
Date: December 19, 2013, 4:51:37 PM EST
To: <mir...@googlegroups.com>
Subject: imdiff: No space left on device
Hi MIRIAD users,�
I�m trying to use imdiff to compare two maps, and when I do I get the following error:
Varied:� amplitude,offset,x-shift,y-shift,expansion### Warning [imdiff]:� Error writing to scratch file; check your $TMPDIR### Fatal Error [imdiff]:� No space left on device
Also a window pops up that says �Low Disk Space �Terminal has detected that the system disk is running low on swap space. Scrollback buffers may automatically be trimmed to conserved memory.��
I�m at a loss because I have over 300 GB free on my hard drive. I can make all sorts of large files in other ways without a problem. I checked to make sure I can write stuff to $TMPDIR (I set $TMPDIR to /tmp and copied some large files to it). I realize that there�s a good chance this isn�t a miriad problem but I�m out of ideas, and my googling skills haven�t turned anything up. Suggestions appreciated!
Best,�
Statia
--
Chat Hull
Graduate Student
Radio Astronomy Laboratory
UC Berkeley
LinkedIn:�linkedin.com/in/chathullWeb:�astro.berkeley.edu/~chat
Reply all
Reply to author
Forward
0 new messages