comerford

Fri Apr 20 16:22:27 [rsMgr] replSet can't see a majority, will not try to elect self
Fri Apr 20 16:22:28 [initandlisten] connection accepted from 192.168.2.71:59427 #3990253 (2 connections now open)
Fri Apr 20 16:22:28 [conn3990253] end connection 192.168.2.71:59427 (1 connection now open)
Fri Apr 20 16:22:28 [initandlisten] connection accepted from 192.168.2.72:55076 #3990254 (2 connections now open)
Fri Apr 20 16:22:28 [conn3990254] end connection 192.168.2.72:55076 (1 connection now open)
Fri Apr 20 16:22:30 [initandlisten] connection accepted from 192.168.2.71:59428 #3990255 (2 connections now open)
Fri Apr 20 16:22:30 [conn3990255] end connection 192.168.2.71:59428 (1 connection now open)
Fri Apr 20 16:22:30 [initandlisten] connection accepted from 192.168.2.72:55077 #3990256 (2 connections now open)
Fri Apr 20 16:22:30 [conn3990256] end connection 192.168.2.72:55077 (1 connection now open)
Fri Apr 20 16:22:32 [initandlisten] connection accepted from 192.168.2.71:59429 #3990257 (2 connections now open)

comerford

// delete oplog, crete new and empty
testReplSet:PRIMARY> use local
switched to db local
testReplSet:PRIMARY> db.oplog.rs.drop()
true
testReplSet:PRIMARY> db.createCollection("oplog.rs", {capped:1, size: 2*1024*1024, autoIndexId:false})
{ "ok" : 1 }
// run the command to get the fields
testReplSet:PRIMARY> db.getReplicationInfo()
{

comerford

// kills long running ops in MongoDB (taking seconds as an arg to define "long")
// attempts to be a bit safer than killing all by excluding replication related operations
// and only targeting queries as opposed to commands etc.
 
killLongRunningOps = function(maxSecsRunning) {
    currOp = db.currentOp();
    for (oper in currOp.inprog) {
        op = currOp.inprog[oper-0];
        if (op.secs_running > maxSecsRunning && op.op == "query" && !op.ns.startsWith("local")) {
            print("Killing opId: " + op.opid

comerford

// start a shell from the command line, do not connect to a database
./mongo --nodb
// using that shell start a new cluster, with a 1MB chunk size
cluster = new ShardingTest({shards: 2, chunksize: 1});
// open another shell (previous one will be full of logging and not actually connected to anything)
./mongo --port 30999
// stop the balancer
sh.stopBalancer()
sh.getBalancerState()
// select test DB, enable sharding

comerford

// function will take a number of days, a collection name, an index Field name, and a boolean as args
// it assumes the index is ObjectID based and creates an ObjectID with a timestamp X days in the past
// Finally, it queries that index/data (controlled by boolean), loading it into memory
// 
// Example - 2 days data, foo collection, _id index, pre-heat index only
// preHeatData(2, "foo", "_id", true)
// Example - 7 days data, foo collection, _id index, pre-heat data also
// preHeatData(7, "foo", "_id", false)
// Example - 2 days data, bar collection, blah index, pre-heat index only
// preHeatData(2, "bar", "blah", false)

comerford

// simple for loop to insert 100k records into the test databases
var testDB = db.getSiblingDB("test");
// drop the collection, avoid dupes
testDB.timecheck.drop();
for(var i = 0; i < 100000; i++){
    testDB.timecheck.insert(
        {_id : i}
        )
};

comerford

Keybase proof

I hereby claim:

• I am comerford on github.
• I am comerford (https://keybase.io/comerford) on keybase.
• I have a public key whose fingerprint is A706 6556 2C81 E177 8932 2D88 9454 794E C192 0598

To claim this, I am signing this object:

comerford

#!/bin/bash
#
# author: Tim "xGhOsTkiLLeRx" Brust
# license: CC BY-NC-SA 4.0
# version: 0.2
# date: 07/10/2014
# description: replace (root) password of squashfs from openELEC
# usage: ./openELEC [password] [device] [hash] [user]
# dependencies: mkpassword (whois), squashfs-tools
#

comerford

// these docs, in 2.6, get bucketed into the 256 bucket (size without header = 240)
// From Object.bsonsize(db.data.findOne()), the size is actually 198 for reference, so add 16 to that for an exact fit
// with that doc size, 80,000 is a nice round number under the 16MiB limit, so will use that for the inner loop
 
// We are shooting for ~16 GiB of data, without indexes, so do 1,024 iterations (512 from each client)
// This will mean being a little short (~500MiB) in terms of target data size, but keeps things simple
 
for(var j = 0; j < 512; j++){ //
  bigDoc = [];
  for(var i = 0; i < 80000; i++){

comerford

storage:
    dbPath: "/ssd/db/wt_snappy"
    engine: "wiredTiger"
systemLog:
    destination: file
    path: "/data/wt_snappy/mongodb.log"
processManagement:
    fork: true