On Apr 04 2016, Kevin Grace <[email protected]> wrote: > S3QL filesystem dismounted on its own. When trying to remount, got error > saying to rerun FSCK. Running FSCK results in the following error > everytime. We have uploaded quiet a bit of data and wondering if we're > hitting a size/object limit,etc.
Sort of. > Cycling metadata backups... > Backing up old metadata... > Encountered ConnectionTimedOut exception (send/recv timeout exceeded), > retrying call to Backend._copy_helper for the 3-th time... [...] I suspect that when S3QL instructs the remote server to make a backup copy of the metadata object, the remote server does not give any sign of life until the copy completes. If the metadata is sufficiently large, or the server sufficiently slow, this will take longer than S3QL is willing to wait and you get the above error. As a workaround, you can increase the tcp timeout (see backend options in the S3QL user's guide). A true solution requires a change to the server. It looks like you are using Swift which is quite simply broken in this regard (see https://doc.s3.amazonaws.com/proposals/copy.html for an explanation of how eg. S3 avoids the problem). 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.
