[MB-7185] resetting the data path to a location which had bad data will cause couchbase server to shut down and not respond ( couch server shuts down first ) Created: 14/Nov/12  Updated: 12/Jun/14  Resolved: 12/Jun/14

Status: Resolved
Project: Couchbase Server
Component/s: ns_server
Affects Version/s: 2.0-beta-2, 2.0, 2.0.1, 2.2.0, 2.1.1, 2.5.0
Fix Version/s: 3.0
Security Level: Public

Type: Bug Priority: Major
Reporter: Thuan Nguyen Assignee: Anil Kumar
Resolution: Incomplete Votes: 0
Labels: None
Remaining Estimate: Not Specified
Time Spent: Not Specified
Original Estimate: Not Specified

Triage: Untriaged
Is this a Regression?: Yes

 Description   
i see these crashes after resetting the data path to /data which had data files from build with a different file format ( or let's say the file could be corrupted )

i am not sure what we can do in this case - keep the server running but show that the status is yellow for that node ?


                                                "Incorrect disk header version"}}}
[error_logger:error,2012-11-14T16:09:17.527,ns_1@127.0.0.1:error_logger<0.5.0>:ale_error_logger_handler:log_msg:76]** Generic server couch_server terminating
** Last message in was {'EXIT',<0.10116.0>,
                           {bad_return_value,
                               {database_disk_version_error,
                                   "Incorrect disk header version"}}}
** When Server state == {server,"/data",
                            {re_pattern,0,0,
                                <<69,82,67,80,91,0,0,0,16,0,0,0,1,0,0,0,0,0,0,
                                  0,0,0,0,0,48,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
                                  0,0,0,0,0,0,0,0,93,0,39,25,77,0,0,0,0,32,
                                  224,255,3,254,255,255,135,254,255,255,7,0,0,
                                  0,0,0,0,0,0,0,0,0,0,0,0,0,0,69,26,84,0,39,0>>},
                            0,"Thu, 15 Nov 2012 00:09:17 GMT"}
** Reason for termination ==




 Comments   
Comment by Farshid Ghods (Inactive) [ 14/Nov/12 ]
i think this is not a blocker or even major becasue we wont have users with earlier builds and we will keep data files compatible between 2.0 versions.

Comment by Farshid Ghods (Inactive) [ 14/Nov/12 ]
manifest :
<manifest>
  <remote name="couchbase" fetch="git://github.com/couchbase/"/>
  <remote name="membase" fetch="git://github.com/membase/"/>
  <remote name="apache" fetch="git://github.com/apache/"/>
  <remote name="erlang" fetch="git://github.com/erlang/"/>

  <default remote="couchbase" revision="master"/>

  <project name="tlm" path="tlm" revision="12abea946eafd7411273d18a10ae1f84390db3d4">
    <copyfile src='Makefile.top' dest='Makefile'/>
  </project>
  <project name="bucket_engine" path="bucket_engine" revision="70b3624abc697b7d18bf3d57f331b7674544e1e7"/>
  <project name="ep-engine" path="ep-engine" revision="a8db69fa759d333246e690c52ef9497c858012ef"/>
  <project name="libconflate" path="libconflate" revision="2cc8eff8e77d497d9f03a30fafaecb85280535d6"/>
  <project name="libmemcached" path="libmemcached" revision="ca739a890349ac36dc79447e37da7caa9ae819f5"
           remote="membase"/>
  <project name="libvbucket" path="libvbucket" revision="00d3763593c116e8e5d97aa0b646c42885727398"/>
  <project name="membase-cli" path="membase-cli" revision="f75b10e3a11e41b28568dde028a4284efdab084d"
           remote="membase"/>
  <project name="memcached" path="memcached" revision="7ea975a93a0231393502af4ca98976eee8a83386"
           remote="membase"/>
  <project name="moxi" path="moxi" revision="52a5fa887bfff0bf719c4ee5f29634dd8707500e"/>
  <project name="ns_server" path="ns_server" revision="56a7e106905000cbbd8d9a20e77d0bfe9c1d2fda"/>
  <project name="portsigar" path="portsigar" revision="1bc865e1622fb93a3fe0d1a4cdf18eb97ed9d600"/>
  <project name="sigar" path="sigar" revision="63a3cd1b316d2d4aa6dd31ce8fc66101b983e0b0"/>
  <project name="couchbase-examples" path="couchbase-examples" revision="c59551adf11860963c1bba028adf33529a4a4d4a"/>
  <project name="couchbase-python-client" path="couchbase-python-client" revision="006c1aa8b76f6bce11109af8a309133b57079c4c"/>
  <project name="couchdb" path="couchdb" revision="00105ea6d44c15aefcf514a9a2c65fbbb6151830"/>
  <project name="couchdbx-app" path="couchdbx-app" revision="a1cbfcaa0091b4b7038bbe949b9d48787cc526eb"/>
  <project name="couchstore" path="couchstore" revision="b5937c4479bf05dcc67264efe19abaf52870a127"/>
  <project name="geocouch" path="geocouch" revision="849d5443689b1924f097548af864c539bffcc929"/>
  <project name="mccouch" path="mccouch" revision="88701cc326bc3dde4ed072bb8441be83adcfb2a5"/>
  <project name="testrunner" path="testrunner" revision="08155ea827ea0508f47406b36e7c892d8591abf9"/>
  <project name="otp" path="otp" revision="b6dc1a844eab061d0a7153d46e7e68296f15a504"
           remote="erlang"/>
  <project name="icu4c" path="icu4c" revision="26359393672c378f41f2103a8699c4357c894be7"
           remote="couchbase"/>
  <project name="snappy" path="snappy" revision="5681dde156e9d07adbeeab79666c9a9d7a10ec95"
           remote="couchbase"/>
  <project name="v8" path="v8" revision="447decb75060a106131ab4de934bcc374648e7f2"
           remote="couchbase"/>
  <project name="gperftools" path="gperftools" revision="8f60ba949fb8576c530ef4be148bff97106ddc59"
           remote="couchbase"/>
  <project name="pysqlite" path="pysqlite" revision="0ff6e32ea05037fddef1eb41a648f2a2141009ea"
           remote="couchbase"/>
</manifest>

version 1950
Comment by Aleksey Kondratenko [ 20/Nov/12 ]
Having any logs here would be actually very helpful
Comment by kzeller [ 06/Dec/12 ]
Added to RN:

If you reset the disk path to a path that contained corrupted data, Couchbase
Server will shut down and non response. Avoid using the same data path if had
contained corrupted data.
Comment by Aleksey Kondratenko [ 04/Sep/13 ]
Before fixing this (or passing to other team for fixing) we need description of what we should do in this case.
Comment by Anil Kumar [ 27/Mar/14 ]
With 3.0 can we retest to find if its still the same.
Comment by Maria McDuff (Inactive) [ 09/May/14 ]
Tony,

can you re-test with latest 3.0 build?
do we have existing (automated) test for this?
Comment by Anil Kumar [ 04/Jun/14 ]
Triage - June 04 2014 Alk, Wayne, Parag, Anil
Comment by Anil Kumar [ 12/Jun/14 ]
Open a new ticket if we see this issue in 3.0.
Generated at Wed Aug 20 09:23:15 CDT 2014 using JIRA 5.2.4#845-sha1:c9f4cc41abe72fb236945343a1f485c2c844dac9.