Seagate / halon

High availability solution
Apache License 2.0
1 stars 0 forks source link

Problem: unpacked m0reportbug artifacts are inconvenient #1551

Closed vvv closed 5 years ago

vvv commented 5 years ago

Unpacked m0reportbug archives cannot be easily attached to Jira tickets. Also navigating them using GitLab UI is a real pain: too many clicks are required to reach any file of interest.

Solution: pass m0reportbug archives to the artifacts unpacked.

vvv commented 5 years ago

merged

vvv commented 5 years ago

enabled an automatic merge when the pipeline for fbe2dc9bf60c7f0f1043e503e1884d17cd16cb06 succeeds

vvv commented 5 years ago

unmarked as a Work In Progress

vvv commented 5 years ago

resolved all discussions

vvv commented 5 years ago

added 7 commits

Compare with previous version

chumakd commented 5 years ago

Right, . is invalid in host names.

centos75_ part comes from m0vg. That will be needed when we have several target platforms.

vvv commented 5 years ago

st-single_halond.m0trace.txt.xz is a long sequence of characters and it is used 3 times. I wanted to put it into a variable. Which would have to be used inside M0VG command and locally. That is why I needed variable expansion within the here document (<<EOF, not <<'EOF').

vvv commented 5 years ago

Where does centos75_ part come from?

As for the dash before {s,ut,m}, I used dot symbol initially, but it would give hostnames like halon6697.s_centos75_cmu.local — those would be invalid AFAIU.

vvv commented 5 years ago

added 2 commits

Compare with previous version

chumakd commented 5 years ago

This yields in VM names like this:

root@qb06n4-m02:/etc# virsh list
 Id    Name                           State
----------------------------------------------------
 1610  halon6697-s_centos75_cmu       running
 1611  halon6697-ut_centos75_cmu      running
 1612  halon6697-m_centos75_cmu       running
 1613  halon6697-m_centos75_ssu1      running
 1614  halon6697-m_centos75_ssu2      running
 1615  halon6697-m_centos75_client1   running

if you're OK with it, I'm OK too.

the old naming was like this halon6697s_centos75_cmu

chumakd commented 5 years ago

Do you think \$ are more explicit and conventional and using <<'EOF' would be too subtle for people to notice it?

vvv commented 5 years ago

added 4 commits

Compare with previous version

vvv commented 5 years ago

added 5 commits

Compare with previous version

vvv commented 5 years ago

OK, I've succeeded in breaking artifacts collection code. :boom:

vvv commented 5 years ago

added 2 commits

Compare with previous version

vvv commented 5 years ago

changed the description

vvv commented 5 years ago

marked as a Work In Progress