Dear all,
I am trying to open a 12Gb byte sgy/su file using seismic unix, and I get the below error message.
/home/subbarao/SU43R3/bin/ximage: ealloc.c: malloc failed (Cannot allocate memory)
Is there a file size limitation in SU? or is the error related to display or something? Do I need to change any settings?
Thank you and look forward to your early response.
Cheers, Subbarao.
Suximage opens a pipe using popen to ximage. It strips the headers and sends the headersless data down the pipes. Evidently it allocated memory. Perhaps suximage can be rewritten using memory mapping (mmap) instead of malloc, but perhaps it will work for you if you kill all other memory hogs. Maybe best to reboot and run it before other processes allocate memory.
On Sep 23, 2022, at 09:31, Subbarao Yelisetti ysrhcu@gmail.com wrote:
Dear all,
I am trying to open a 12Gb byte sgy/su file using seismic unix, and I get the below error message.
/home/subbarao/SU43R3/bin/ximage: ealloc.c: malloc failed (Cannot allocate memory)
Is there a file size limitation in SU? or is the error related to display or something? Do I need to change any settings?
Thank you and look forward to your early response.
Cheers, Subbarao.
--
Dr. Subbarao Yelisetti Associate Professor of Geophysics Graduate Coordinator for Petrophysics Department of Physics and Geosciences Texas A&M University - Kingsville Hill Hall 207 MSC 175, 700 University Blvd Kingsville, Texas 78363-8202 Phone: 361-593-4894 web: https://www.tamuk.edu/artsci/departments/phge/faculty/yelisetti.html _______________________________________________ Seisunix mailing list Seisunix@mailman.seismic-unix.org https://mailman.seismic-unix.org/listinfo/seisunix
Hi, any computer is finite, so, there is limitation to everything... What is the available memory size of your computer ?
12 Gb is quite a large number of samples for a single screen, even 4K. The same remark applies to human eyes. Did you try to reduce the size of the data ?
D.
Le 23/09/2022 à 18:28, Subbarao Yelisetti a écrit :
Dear all,
I am trying to open a 12Gb byte sgy/su file using seismic unix, and I get the below error message.
/home/subbarao/SU43R3/bin/ximage: ealloc.c: malloc failed (Cannot allocate memory)
Is there a file size limitation in SU? or is the error related to display or something? Do I need to change any settings?
Thank you and look forward to your early response.
Cheers, Subbarao.
--
Dr. Subbarao Yelisetti Associate Professor of Geophysics Graduate Coordinator for Petrophysics Department of Physics and Geosciences Texas A&M University - Kingsville Hill Hall 207 MSC 175, 700 University Blvd Kingsville, Texas 78363-8202 Phone: 361-593-4894 <tel:361-593-4894> web: https://www.tamuk.edu/artsci/departments/phge/faculty/yelisetti.html _______________________________________________ Seisunix mailing list Seisunix@mailman.seismic-unix.org https://mailman.seismic-unix.org/listinfo/seisunix
Greetings. Obviously, attempting to plot a 12Gb file on the screen is not a good idea, considering the number of available pixels on your screen is no more than a few thousand.
Try instead $ suwind < your_file.su count=2000 | suximage perc=99 &
$ suwind < your_file.su count=2000 skip=2000 | suximage perc=99 &
This sort of thing.
John Stockwell Visit: https://wiki.Seismic-Unix.org The new home of Seismic Un*x
On Sep 23, 2022, at 11:20 AM, Dominique ROUSSET dominique.rousset@univ-pau.fr wrote: Dear Hi, any computer is finite, so, there is limitation to everything... What is the available memory size of your computer ?
12 Gb is quite a large number of samples for a single screen, even 4K. The same remark applies to human eyes. Did you try to reduce the size of the data ?
D.
Le 23/09/2022 à 18:28, Subbarao Yelisetti a écrit :
Dear all,
I am trying to open a 12Gb byte sgy/su file using seismic unix, and I get the below error message.
/home/subbarao/SU43R3/bin/ximage: ealloc.c: malloc failed (Cannot allocate memory)
Is there a file size limitation in SU? or is the error related to display or something? Do I need to change any settings?
Thank you and look forward to your early response.
Cheers, Subbarao.
--
Dr. Subbarao Yelisetti Associate Professor of Geophysics Graduate Coordinator for Petrophysics Department of Physics and Geosciences Texas A&M University - Kingsville Hill Hall 207 MSC 175, 700 University Blvd Kingsville, Texas 78363-8202 Phone: 361-593-4894 web: https://www.tamuk.edu/artsci/departments/phge/faculty/yelisetti.html _______________________________________________ Seisunix mailing list Seisunix@mailman.seismic-unix.org https://mailman.seismic-unix.org/listinfo/seisunix-- |Dominique Rousset LFCR UMR5150 / E2S UPPA / Dpt des géosciences| |Univ. Pau et des Pays de l'Adour BP 1155 F-64013 Pau Cedex - France| |mailto:dominique.rousset@univ-pau.fr http://lfc.univ-pau.fr%7C | Off: +33 559 407 423 Cell: +33 660 861 391 | _______________________________________________ Seisunix mailing list Seisunix@mailman.seismic-unix.org https://mailman.seismic-unix.org/listinfo/seisunix
Well, it is free software and whoever wish to use it can do his or hers upgrades to achieve the goals ! Sam T.
Sent from Yahoo Mail for iPhone
On Friday, September 23, 2022, 4:04 PM, Don Habib hbadro_19@hotmail.com wrote:
Hi everyone, it could not allocate enough memory to open the file, and. it's because of the program, I think it's limited. Regards
Sent from my Huawei phone
-------- Original Message -------- Subject: [Seisunix] suximage error From: Subbarao Yelisetti To: seisunix@mailman.seismic-unix.org CC:
Dear all, I am trying to open a 12Gb byte sgy/su file using seismic unix, and I get the below error message. /home/subbarao/SU43R3/bin/ximage: ealloc.c: malloc failed (Cannot allocate memory)
Is there a file size limitation in SU? or is the error related to display or something? Do I need to change any settings? Thank you and look forward to your early response. Cheers,Subbarao.
On Fri, 23 Sep 2022 22:03:34 +0100, Don Habib hbadro_19@hotmail.com wrote:
Hi everyone, it could not allocate enough memory to open the file, and. it's because of the program, I think it's limited.
Hi Habib.
Perhaps it is not just that easy, even because you didn't described the platform you are using.
For example, you have a 12GiB data. Due to the limitation of the SegY/SU format, you are limited to 32Ki data samples, even if the "ns" field in header is defined as "unsigned short", that would allow for 65Ki data samples.
The suximage does not plot the data, it just prepare the data for the ximage program. So, suximage will allocate some memory and ximage will allocate as well.
Besides this the Xwindow libraries will allocate some memory for its structure.
As you gave us no information about your computer, lets say that your computer has 8GiB of RAM and other 8GiB of swap. When we thing of storing 12GiB of data in memory it will not only allocate 12GiB. I don't now data characteristics, so I can only guess. Lets say that 10% of those 12GiB are headers that are not stored for plot, it leaves about 10GiB just for the data, we are not talking about Xlib structures, system used memory for the kernel and all services, the structure used by your desktop itself, the memory space used for (say) browser and so on.
Following with the guesswork, as no information beyond data size was informed. In my computer right now I am navigating internet and answering this message. In this state the computer is allocating about 7.8GiB among programs, services and buffers. We can take out buffer and cached data, which would reduce the memory in use to about 5.6GiB.
If you add 5.6Gib with 10GiB it will produce almost the 16GiB we guessed your computer could have summing RAM and SWAP. There is an amount that is not showed that is that used by the kernel itself. If adding all of it we reach more that the available total memory (RAM + SWAP) the system will not allow the allocation. And the whole SU package has nothing to do with it.
To be honest, I have about 40 years in geophysics work, most of it in seismic data processing. Even when we thing about plotting 3D data in interpretation stations, never heard of anyone trying to plot 12GiB of data. To me, this information alone, is kind of mind boggling on how do you intent to visualize this amount of data in a computer terminal.
So, state that the fault is of the software without a single information of the data (pre/post stack?) and what are you intending to visualize, and without any information about the computer you are using, even if there is any problem with the software, where do you think anyone would start to tackle this pseudo problem?
Perhaps if you provided more information, someone could try to help you.
HTH
Roxo
I suggest you use CSM supported seismic data viewer SEAVIEW. I had no issues with file sizes with seaview.
Necati Gülünay n.gulunay@protonmail.com +90 541 842 0069
Web. https://www.seisdatapro.com
Sent from Proton Mail for iOS
On Sat, Sep 24, 2022 at 02:32, Fernando M. Roxo da Motta petro@roxo.org wrote:
On Fri, 23 Sep 2022 22:03:34 +0100, Don Habib hbadro_19@hotmail.com wrote:
Hi everyone, it could not allocate enough memory to open the file, and. it's because of the program, I think it's limited.
Hi Habib.
Perhaps it is not just that easy, even because you didn't described the platform you are using.
For example, you have a 12GiB data. Due to the limitation of the SegY/SU format, you are limited to 32Ki data samples, even if the "ns" field in header is defined as "unsigned short", that would allow for 65Ki data samples.
The suximage does not plot the data, it just prepare the data for the ximage program. So, suximage will allocate some memory and ximage will allocate as well.
Besides this the Xwindow libraries will allocate some memory for its structure.
As you gave us no information about your computer, lets say that your computer has 8GiB of RAM and other 8GiB of swap. When we thing of storing 12GiB of data in memory it will not only allocate 12GiB. I don't now data characteristics, so I can only guess. Lets say that 10% of those 12GiB are headers that are not stored for plot, it leaves about 10GiB just for the data, we are not talking about Xlib structures, system used memory for the kernel and all services, the structure used by your desktop itself, the memory space used for (say) browser and so on.
Following with the guesswork, as no information beyond data size was informed. In my computer right now I am navigating internet and answering this message. In this state the computer is allocating about 7.8GiB among programs, services and buffers. We can take out buffer and cached data, which would reduce the memory in use to about 5.6GiB.
If you add 5.6Gib with 10GiB it will produce almost the 16GiB we guessed your computer could have summing RAM and SWAP. There is an amount that is not showed that is that used by the kernel itself. If adding all of it we reach more that the available total memory (RAM + SWAP) the system will not allow the allocation. And the whole SU package has nothing to do with it.
To be honest, I have about 40 years in geophysics work, most of it in seismic data processing. Even when we thing about plotting 3D data in interpretation stations, never heard of anyone trying to plot 12GiB of data. To me, this information alone, is kind of mind boggling on how do you intent to visualize this amount of data in a computer terminal.
So, state that the fault is of the software without a single information of the data (pre/post stack?) and what are you intending to visualize, and without any information about the computer you are using, even if there is any problem with the software, where do you think anyone would start to tackle this pseudo problem?
Perhaps if you provided more information, someone could try to help you.
HTH
Roxo
-- ---------------- Non luctari, ludare -------------------+ WYSIWYG Fernando M. Roxo da Motta petro@roxo.org | Editor? Except where explicitly stated I speak on my own behalf.| VI !! PU5RXO | PX5Q6048 | I see text, ------------ Quis custodiet ipsos custodes?-------------+ I get text!
Seisunix mailing list Seisunix@mailman.seismic-unix.org https://mailman.seismic-unix.org/listinfo/seisunix
John’s suggestion to use suwind to reduce the number of traces us good and his point that you anyway don’t have more than a couple of thousands pixels across the screen. However, i myself often display more traces than pixels across and more time samples than pixels down and then i rubber band to zoom up on interesting time-offset windows. Interactive windowing is reversible and better than irreversible command line windowing. But of course at some size you hit the memory limits. Unix’ vmstat can be useful to see if there are other memory hogs that you may want to kill and rob their virtual memory.
Suximage could be written with mmap instead of malloc but this may be a big change for a questionable value added.
On Sep 23, 2022, at 16:33, Fernando M. Roxo da Motta petro@roxo.org wrote:
On Fri, 23 Sep 2022 22:03:34 +0100, Don Habib hbadro_19@hotmail.com wrote:
Hi everyone, it could not allocate enough memory to open the file, and. it's because of the program, I think it's limited.
Hi Habib.
Perhaps it is not just that easy, even because you didn't described the platform you are using.
For example, you have a 12GiB data. Due to the limitation of the SegY/SU format, you are limited to 32Ki data samples, even if the "ns" field in header is defined as "unsigned short", that would allow for 65Ki data samples.
The suximage does not plot the data, it just prepare the data for the ximage program. So, suximage will allocate some memory and ximage will allocate as well.
Besides this the Xwindow libraries will allocate some memory for its structure.
As you gave us no information about your computer, lets say that your computer has 8GiB of RAM and other 8GiB of swap. When we thing of storing 12GiB of data in memory it will not only allocate 12GiB. I don't now data characteristics, so I can only guess. Lets say that 10% of those 12GiB are headers that are not stored for plot, it leaves about 10GiB just for the data, we are not talking about Xlib structures, system used memory for the kernel and all services, the structure used by your desktop itself, the memory space used for (say) browser and so on.
Following with the guesswork, as no information beyond data size was informed. In my computer right now I am navigating internet and answering this message. In this state the computer is allocating about 7.8GiB among programs, services and buffers. We can take out buffer and cached data, which would reduce the memory in use to about 5.6GiB.
If you add 5.6Gib with 10GiB it will produce almost the 16GiB we guessed your computer could have summing RAM and SWAP. There is an amount that is not showed that is that used by the kernel itself. If adding all of it we reach more that the available total memory (RAM + SWAP) the system will not allow the allocation. And the whole SU package has nothing to do with it.
To be honest, I have about 40 years in geophysics work, most of it in seismic data processing. Even when we thing about plotting 3D data in interpretation stations, never heard of anyone trying to plot 12GiB of data. To me, this information alone, is kind of mind boggling on how do you intent to visualize this amount of data in a computer terminal.
So, state that the fault is of the software without a single information of the data (pre/post stack?) and what are you intending to visualize, and without any information about the computer you are using, even if there is any problem with the software, where do you think anyone would start to tackle this pseudo problem?
Perhaps if you provided more information, someone could try to help you.
HTH
Roxo
-- ---------------- Non luctari, ludare -------------------+ WYSIWYG Fernando M. Roxo da Motta petro@roxo.org | Editor? Except where explicitly stated I speak on my own behalf.| VI !! PU5RXO | PX5Q6048 | I see text, ------------ Quis custodiet ipsos custodes?-------------+ I get text!
Seisunix mailing list Seisunix@mailman.seismic-unix.org https://mailman.seismic-unix.org/listinfo/seisunix
seisunix@mailman.seismic-unix.org
-
Dominique ROUSSET -
Don Habib -
Fernando M. Roxo da Motta -
John Stockwell -
N.Gülünay -
Sam Sampanthan -
Shuki Ronen -
Subbarao Yelisetti