Skip to main content
Question

Vertica Node would not start - LOST CONTACT

  • Jun 12, 2025 5:41 PM UTC
  • 2 replies
  • 17 views
dimitri_p
Forum|alt.badge.img+2

On Vertica 12 one of the nodes went down and when trying to re-start vertica on the host we get LOST CONTACT in Admintools.

We get this in Vertica.log (and no other errors):

2025-06-11 20:28:45.303 RecoverTable:0x7feb1effd700-16000000000093f [Txn] <INFO> dropping partitioned StorageContainer (ROS): 0x16000001ae9658d
2025-06-11 20:28:45.306 EEThread:0x7fe7cf7fe700-16000000000091e [Main] <PANIC> Received fatal signal SIGSEGV.
2025-06-11 20:28:45.306 EEThread:0x7fe7cf7fe700-16000000000091e [Main] <PANIC> Info: si_code: 1, si_pid: 1183312701, si_uid: 0, si_addr: 0x4687eb3d

and this in ErrorReport.txt

Restart Vertica on host and Restart Cluster do not help, we always end up with one node down.

Any idea what else to try?

    2 replies

    SruthiA
    Forum|alt.badge.img+1
    • SruthiA
    • Participating Frequently
    • Jun 12, 2025 10:27 PM UTC

    please raise a support case. it may be a bug.

    dimitri_p
    Forum|alt.badge.img+2
    • dimitri_p
    • Author
    • Participating Frequently
    • Jun 24, 2025 11:57 AM UTC

    In case you've found this post using Search and are having the same issue, we've fixed ours by formatting the data filesystem on the problematic node and having the node recovered by the rest of the cluster (restart Vertica on Host). So it looks like it was some sort of metadata corruption that vertica couldn't handle automatically.

    Still scrolling?
    Ask a question
    Terms & ConditionsAccessibility statement