-
Notifications
You must be signed in to change notification settings - Fork 3.9k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
roachtest: release-20.1: double init due to restart of node 1 #51965
Comments
irfansharif
added
C-bug
Code not up to spec/doc, specs & docs deemed correct. Solution expected to change code/behavior.
A-testing
Testing tools and infrastructure
labels
Jul 28, 2020
irfansharif
changed the title
roachtest: double init due to restart of node 1
roachtest: release-20.1: double init due to restart of node 1
Jul 28, 2020
Note to self: This will affect release-20.1 roachtests going forward, so close them out as well when addressing this. This is also another reason to have #51897, it's a bit surprising that CI lets you land roach{prod,test} changes without testing how they affect tests against 20.1 and earlier. Blocking #51919. |
This was referenced Jul 28, 2020
irfansharif
added a commit
to irfansharif/cockroach
that referenced
this issue
Jul 28, 2020
Fixes cockroachdb#51965 (and all referencing issues). Roachprod clusters running v20.1+ crdb nodes persist this `cluster-bootstrapped` file on disk after explicitly bootstrapping the cluster. Roachprod then uses the existence of this file to avoid doubly bootstrapping the cluster. Given cockroachdb#51897 remains unresolved, master-built roachprod is used to run roachtests against the 20.1 branch. Some of those roachtests test mixed-version clusters that start off at 19.2. Consequently, we manually add this file where roachprod expects to find it for already-initialized clusters. (This is a pretty gross hack, that we should address by addressing cockroachdb#51897.) Release note: None
craig bot
pushed a commit
that referenced
this issue
Jul 29, 2020
52040: roachtest: fix release-20.1 roachtests failing due to double-init r=RaduBerinde a=irfansharif Fixes #51965 (and all referencing issues). Roachprod clusters running v20.1+ crdb nodes persist this `cluster-bootstrapped` file on disk after explicitly bootstrapping the cluster. Roachprod then uses the existence of this file to avoid doubly bootstrapping the cluster. Given #51897 remains unresolved, master-built roachprod is used to run roachtests against the 20.1 branch. Some of those roachtests test mixed-version clusters that start off at 19.2. Consequently, we manually add this file where roachprod expects to find it for already-initialized clusters. (This is a pretty gross hack, that we should address by addressing #51897.) Release note: None Co-authored-by: irfan sharif <[email protected]>
craig bot
pushed a commit
that referenced
this issue
Jul 29, 2020
52015: sql: use a structured error to detect roachpb.BatchTimestampBeforeGCE… r=ajwerner a=ajwerner …rror Fixes #50198. See the issue for more details. This issue is so obscure it does not deserve a release note. Release note: None 52040: roachtest: fix release-20.1 roachtests failing due to double-init r=irfansharif a=irfansharif Fixes #51965 (and all referencing issues). Roachprod clusters running v20.1+ crdb nodes persist this `cluster-bootstrapped` file on disk after explicitly bootstrapping the cluster. Roachprod then uses the existence of this file to avoid doubly bootstrapping the cluster. Given #51897 remains unresolved, master-built roachprod is used to run roachtests against the 20.1 branch. Some of those roachtests test mixed-version clusters that start off at 19.2. Consequently, we manually add this file where roachprod expects to find it for already-initialized clusters. (This is a pretty gross hack, that we should address by addressing #51897.) Release note: None Co-authored-by: Andrew Werner <[email protected]> Co-authored-by: irfan sharif <[email protected]>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
This is the tracking issue for the set of failures in provisional_202007271721_v20.1.4
#51952
#51953
#51954
#51955
#51956
#51957
#51958
#51960
They all fail with "has already been initialized" errors, and are all attributable to recent changes made to
roachprod
starting from #51329, where we changed it to explicitly initialize the cluster when starting node 1, when there's nocluster-bootstrapped
file found on disk. For all these failing tests we seem to always be starting node 1 mid-test, where upon not finding thecluster-bootstrapped
file on disk it expectedly failing to initialize the cluster and thus errors out. The tests that fail in this pattern are "mixed-version" tests, and given we're targeting the release-20.1 branch here, they all start off with nodes running 19.2 (never actually persisting thecluster-bootstrapped
file on disk during bootstrap, which only happens when initializing a 20.1+ cluster).The text was updated successfully, but these errors were encountered: