On Feb 28 2017, Riku Bister <[email protected]> wrote:
> Whatever application you are using to provide the "local" file system
>> sometimes gives an "Bad Address" error when S3QL tries to access
>> it. This problem needs to be fixed there.
>
> Is there any way to get retries with timeout instead a fs crash on first
> error?
As far as the backend directory is concerned, S3QL is a regular
application like cp, unzip, or libreoffice.
Would you expect those applications to add a special check so that "if
the filesystem is ACD, and there is an unexpected error when trying to
read data, just wait and try again"?
I hope you would not. For the same reasons, S3QL will not have any such
code. Just because S3QL happens to provide a file system does not mean
that you can put it on top of another filesystem and expect it to
magically work around the lower layer's brokenness.
> i could do a automate script to remount, but fsck takes way too long
> for doing it more than half hour per fs
In that case, you should not run S3QL with a backend that you know to be
broken.
Best,
-Nikolaus
--
GPG encrypted emails preferred. Key id: 0xD113FCAC3C4E599F
Fingerprint: ED31 791B 2C5C 1613 AF38 8B8A D113 FCAC 3C4E 599F
»Time flies like an arrow, fruit flies like a Banana.«
--
You received this message because you are subscribed to the Google Groups
"s3ql" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
For more options, visit https://groups.google.com/d/optout.
