[SERVER-2670] Assertion failure _a != -1 db/pdfile.h 425 Created: 03/Mar/11 Updated: 30/Mar/12 Resolved: 17/Mar/11 |
|
| Status: | Closed |
| Project: | Core Server |
| Component/s: | None |
| Affects Version/s: | 1.8.0-rc0 |
| Fix Version/s: | None |
| Type: | Bug | Priority: | Major - P3 |
| Reporter: | David Mytton | Assignee: | Eliot Horowitz (Inactive) |
| Resolution: | Cannot Reproduce | Votes: | 0 |
| Labels: | None | ||
| Remaining Estimate: | Not Specified | ||
| Time Spent: | Not Specified | ||
| Original Estimate: | Not Specified | ||
| Attachments: |
|
| Operating System: | ALL |
| Participants: |
| Description |
|
Running a count on a newly created index caused an assertion. mongod then complained the db was corrupt and to run a repair, which I did, which resolved the issue. pb1:PRIMARY> use postbacks ) // 2GB { "ok" : 1 }pb1:PRIMARY> db.postbacks.ensureIndex( { '_id' : 1}, {background: true}) , {background: true}) , ) ) ) , , {background: true}) pb1:PRIMARY> db.postbacks.stats() }, , {background: true}) , {background: true}) , ) |
| Comments |
| Comment by Eliot Horowitz (Inactive) [ 04/Mar/11 ] |
|
There were a number of fixes that might be related in 1.8.0-rc1 going out shortly. |
| Comment by David Mytton [ 04/Mar/11 ] |
|
This was a new database on a new mongo instance. Dropping/re-indexing works fine without --dur enabled but with --dur it asserts. |
| Comment by Eliot Horowitz (Inactive) [ 04/Mar/11 ] |
|
Is there any chance this db was corrupt before creating the index? |
| Comment by David Mytton [ 03/Mar/11 ] |
|
Done. david@pb1b /mongodbdata/dump/dump/postbacks: scp -P 722 postbacks.bson david@www.10gen.com: |
| Comment by Eliot Horowitz (Inactive) [ 03/Mar/11 ] |
|
Sorry, should be scp -P 722 yourfile david@www.10gen.com: don't put anything after the colon |
| Comment by David Mytton [ 03/Mar/11 ] |
|
david@pb1b /mongodbdata/dump/dump/postbacks: scp -v -P 722 david@www.10gen.com: postbacks.bson |
| Comment by Eliot Horowitz (Inactive) [ 03/Mar/11 ] |
|
Correct, for the public its write only. |
| Comment by David Mytton [ 03/Mar/11 ] |
|
Will do. This isn't publicly accessible? |
| Comment by Eliot Horowitz (Inactive) [ 03/Mar/11 ] |
|
Can you scp to us? scp -P 722 david@www.10gen.com: |
| Comment by David Mytton [ 03/Mar/11 ] |
|
It's a couple of GB: david@pb1b /mongodbdata/dump: ls -alh dump/postbacks/ |
| Comment by Eliot Horowitz (Inactive) [ 03/Mar/11 ] |
|
Is this is a small collection? |
| Comment by David Mytton [ 03/Mar/11 ] |
|
Actually scrap that, I can reproduce with both background and not. It's durability that makes the difference - enabled = crash. not enabled = no crash. |
| Comment by David Mytton [ 03/Mar/11 ] |
|
Seems to be related to using a background index - I can reproduce if I use background, but not if it's a regular blocking index creation. |
| Comment by David Mytton [ 03/Mar/11 ] |
|
It's either true or false. The whole doc itself is quite big though. |
| Comment by Eliot Horowitz (Inactive) [ 03/Mar/11 ] |
|
What kind of field is "p"? |
| Comment by David Mytton [ 03/Mar/11 ] |
|
Thu Mar 3 00:28:17 [conn214320] query postbacks.$cmd ntoreturn:1 command: { create: "postbacks", capped: true, size: 2147483648.0, max: undefined } reslen:53 14390ms } bytes:139 nreturned:1 19197ms for postbacks.postbacks for postbacks.postbacks background reslen:284 433ms reslen:224 1283ms failed with exception: exception: 0 assertion db/pdfile.h:425 |