Tired. Exhausted. Smelly

Geekery — December 7, 2008 at 5:47 pm

Ended my last primary on-call shift of the year just now, and not a minute too soon. I like to say that my job is at the intersection of system administration and adrenaline because for 37 hours a weekend we are constantly solving issues where data loss is not an option and speed is everything.

Anyway, two choice quotes from my weekend, emphasis mine:

From a Sybase ASE error log:

2008/12/07 02:34:34.40 server Error: 9970, Severity: 20, State: 1 2008/12/07 02:34:34.40 server DBCC cannot update the finish time in dbcc_operation_log table for this operation(opid = ’182′) of database ‘sybsystemprocs’. This can be patched by executing sp_dbcc_patch_finishtime.

I love software that has been patched to tell you how to actually fix the issue but won’t actually fix it for you.

Next up, an email from a first-level system administrator:

We see that the host is now responding. But host is on high load. Please check your applications. piias1332 /ms/user/v/venusrin 63$ rup pisdb38 pisdb38 up 29 days, 2:42, load average: 1.77 1.73 1.62

Um. No. Just no.

2 Comments »

  1. why no?

    Comment by eric schmidt — August 11, 2009 @ 12:37 am
  2. Load average on Solaris is defined as the number of running processes plus the number of runnable processes. (On Linux, it includes the number of outstanding I/O requests). A load average of 1 means there is one process that can be run for the duration of the sample. Having one process around is not a high load.

    Comment by Eric — August 11, 2009 @ 4:25 pm

RSS feed for comments on this post. TrackBack URI

Leave a comment

This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 2.5 License. | Eric Garrido