My apologies...I have obviously confused two different threads of discussion.

One thread, from several weeks ago, was regarding file corruption when a wholly 
different code base is producing HDF5 bytes-on-disk and then a subsequent 
read/write by an HDF5 tool was corrupting the file.

This thread, from last week, was regarding file corruption due to 
crash/shutdown before H5Fclose.

Again, sorry for confusion.

Mark


"Hdf-forum on behalf of Dana Robinson" wrote:

Hi all,

(I'm jumping in to the middle of this discussion while still on vacation, so 
please excuse me if I'm missing something.)

Those documents describe cache flushing. Flush ordering should not affect the 
file format – any transient 'file corruption' would be due to a second reader 
inspecting an incomplete file, which techniques like SWMR are designed to 
address.

Incompatibilities between the official HDF5 library and third-party HDF5 
libraries should probably be considered bugs in one or the other (or maybe even 
both!), as long as they are truly holding to the published HDF5 file format.

Dana

From: Hdf-forum <[email protected]> on behalf of "Miller, 
Mark C." <[email protected]>
Reply-To: HDF List <[email protected]>
Date: Monday, September 25, 2017 at 10:52
To: HDF List <[email protected]>
Subject: Re: [Hdf-forum] Avoiding corruption of the HDF5 File

Hi Quincey,

One question though...Is it possible to produce bytes-on-disk format from a 
wholly different code base that is nonetheless compatible with HDF5 proper?

Your answer seems to suggest it is NOT possible without using (some of) the 
HDF5 code base.

That would be a shame as it suggests there is no longer a well defined 
bytes-on-disk format apart from whatever the HDF5 implementation produces.

Mark


"Hdf-forum on behalf of Quincey Koziol" wrote:

Hi Ewan,
There’s two things you can be doing to address file corruption issues:

- For the near term, use the techniques and code for managing the metadata 
cache described here:  
https://support.hdfgroup.org/HDF5/docNewFeatures/FineTuneMDC/RFC%20H5Ocork%20v5%20new%20fxn%20names.pdf

- In the next year or so, we will be finishing the “SWMR” feature, described 
here:  https://support.hdfgroup.org/HDF5/docNewFeatures/NewFeaturesSwmrDocs.html

The metadata cache techniques are rather unsubtle, but will avoid corrupted 
files until the “full” SWMR feature is finished.

Quincey


On Sep 21, 2017, at 8:33 PM, Ewan Makepeace 
<[email protected]<mailto:[email protected]>> wrote:

Dear Experts,

We are building a data acquisition and processing system on top of an HDF5 file 
store. Generally we have been very pleased with HDF5 - great flexibility in 
data structure, performant, small file size, availability of third party data 
access tools etc.

However our system needs to run for 36-48 hours at a time - and we are finding 
that if we (deliberately or accidentally) stop the process while running (and 
writing data) the file is corrupted and we lose all our work.

We are in C# and wrote our access routines on top of HDF5.net<http://HDF5.net> 
(which I understand is deprecated). We tend to keep all active pointer objects 
open for the duration of the process that reads or writes them (file, group and 
dataset handles in particular).

1) Is there a full featured replacement for HDF5.net<http://HDF5.net> now, that 
I was unaware of? Previous contenders were found to be missing support for 
features we depend on. If so will it address the corruption issue?

2) Should we be opening and closing all the entities on every write? I would 
have thought that would dramatically slow access but perhaps not. Guidance?

3) Are there any other tips to making the file less susceptible to corruption 
if writing is abandoned unexpectedly?

Please help - this issue could be serious enough to make us reconsider our 
storage choice, which would be expensive now.

rgds,
Ewan
_______________________________________________
Hdf-forum is for HDF software users discussion.
[email protected]<mailto:[email protected]>
http://lists.hdfgroup.org/mailman/listinfo/hdf-forum_lists.hdfgroup.org
Twitter: https://twitter.com/hdf5

_______________________________________________
Hdf-forum is for HDF software users discussion.
[email protected]
http://lists.hdfgroup.org/mailman/listinfo/hdf-forum_lists.hdfgroup.org
Twitter: https://twitter.com/hdf5

Reply via email to