#2202 new defect

ERROR: UnrecoverableFileError(no recoverable versions)

Reported by: gennylee Owned by: daira
Priority: major Milestone: undecided
Component: unknown Version: 1.10.0
Keywords: transparency usability erasure-coding Cc:
Launchpad Bug:

Description (last modified by zooko)

I configured tahoe-lafs with 5 node, and I configured shares.needed=2, shares.happy=3 and shares.total=5 in the tahoe-client. I turned off all but one my storage's server when i restarted my server, my tahoe-storage configuration say:

ERROR: UnrecoverableFileError(no recoverable versions)
"[Failure instance: Traceback: <class 'allmydata.mutable.common.UnrecoverableFileError'>: no recoverable versions"
/usr/lib/python2.7/dist-packages/twisted/internet/base.py:800:runUntilCurrent
/usr/lib/python2.7/dist-packages/foolscap/eventual.py:26:_turn
/usr/lib/python2.7/dist-packages/twisted/internet/defer.py:362:callback
/usr/lib/python2.7/dist-packages/twisted/internet/defer.py:458:_startRunCallbacks
--- <exception caught here> ---
/usr/lib/python2.7/dist-packages/twisted/internet/defer.py:545:_runCallbacks
/root/allmydata-tahoe-1.10.0/src/allmydata/mutable/filenode.py:400:_get_version

How can I resolve this? Do it possible that I error to write the configuration? Thanks

Attachments (1)

tahoe-grid-status.png (144.1 KB) - added by gennylee at 2014-03-19T09:14:53Z.

Download all attachments as: .zip

Change History (10)

comment:1 Changed at 2014-03-18T16:48:00Z by zooko

  • Description modified (diff)

comment:2 Changed at 2014-03-18T16:49:00Z by zooko

  • Keywords transparency usability erasure-coding added

comment:3 follow-up: Changed at 2014-03-18T16:49:50Z by zooko

Since you set shares.needed=2, then you have to have at least two storage servers running in order to recover files. What did you *think* would happen when you turned off all but one of your servers?

comment:4 in reply to: ↑ 3 ; follow-up: Changed at 2014-03-18T17:10:59Z by gennylee

Replying to zooko:

Since you set shares.needed=2, then you have to have at least two storage servers running in order to recover files. What did you *think* would happen when you turned off all but one of your servers?

I set needed=2 in order to recover files when just 2 storage servers are up, so I'm OK with the idea that with 1 only storage server up I cannot recover any file. I'd expect that when all the storage servers are turned on again, i can again retrieve stored files. The problem is that, instead, the system stays in an unrecoverable state.

ERROR: UnrecoverableFileError(no recoverable versions)
"[Failure instance: Traceback: <class 'allmydata.mutable.common.UnrecoverableFileError'>: no recoverable versions"
/usr/lib/python2.7/dist-packages/twisted/internet/base.py:800:runUntilCurrent
/usr/lib/python2.7/dist-packages/foolscap/eventual.py:26:_turn
/usr/lib/python2.7/dist-packages/twisted/internet/defer.py:362:callback
/usr/lib/python2.7/dist-packages/twisted/internet/defer.py:458:_startRunCallbacks
--- <exception caught here> ---
/usr/lib/python2.7/dist-packages/twisted/internet/defer.py:545:_runCallbacks
/root/allmydata-tahoe-1.10.0/src/allmydata/mutable/filenode.py:400:_get_version

Thanks

Last edited at 2014-03-18T23:59:13Z by zooko (previous) (diff)

comment:5 in reply to: ↑ 4 Changed at 2014-03-19T00:03:09Z by zooko

Replying to gennylee:

I'd expect that when all the storage servers are turned on again, i can again retrieve stored files. The problem is that, instead, the system stays in an unrecoverable state.

Yes, I agree that you should be able to recover the files after you turn all the servers back on! Please tell me what the "Welcome Page" says about how many servers are connected, and please tell me the results of running a "deep check" on that file.

Thanks!

Changed at 2014-03-19T09:14:53Z by gennylee

comment:6 Changed at 2014-03-19T09:20:44Z by gennylee

The "Welcome Page" says that two servers are connected: source:/trac/tahoe-lafs/attachment/ticket/2202/tahoe-grid-status.png The results of running deep check: http://pastebin.com/tiMCcEgw Thanks!

Version 0, edited at 2014-03-19T09:20:44Z by gennylee (next)

comment:7 Changed at 2014-03-19T15:43:29Z by zooko

Here is the content of http://pastebin.com/tiMCcEgw (for the record)

root@storage1:~/allmydata-tahoe-1.10.0/bin# ./tahoe deep-check
ERROR: UnrecoverableFileError(no recoverable versions)
"[Failure instance: Traceback: <class 'allmydata.mutable.common.UnrecoverableFileError'>: no recoverable versions"
/usr/lib/python2.7/dist-packages/twisted/internet/base.py:800:runUntilCurrent
/usr/lib/python2.7/dist-packages/foolscap/eventual.py:26:_turn
/usr/lib/python2.7/dist-packages/twisted/internet/defer.py:362:callback
/usr/lib/python2.7/dist-packages/twisted/internet/defer.py:458:_startRunCallbacks
--- <exception caught here> ---
/usr/lib/python2.7/dist-packages/twisted/internet/defer.py:545:_runCallbacks
/root/allmydata-tahoe-1.10.0/src/allmydata/mutable/filenode.py:400:_get_version

comment:8 Changed at 2014-03-20T21:24:09Z by daira

Well that's horribly bad error reporting :-(

comment:9 Changed at 2014-03-20T21:29:01Z by daira

Error reporting issue filed as #2203.

Note: See TracTickets for help on using tickets.