Mercurial > p > roundup > code
view tools/load_tracker.py @ 6565:2c2dbfc332ba
Try to handle multiple connections better.
The session database is a hot spot. When multiple requests (e.g. 20)
come in at the same time session database contention can get great.
The original code didn't retry session database access when the open
failed. This resulted in errors at the client.
The second pass delayed 0.01 seconds and retried. It was better but we
still had multiple second stalls. I think the first request got in,
everybody else backed up and then retried at the same time. Again they
stepped on each other. With logging I would see many counters go all
the way to low single digits or to -1 indicating falure.
This pass uses randomint to generate delays from 0-.125 seconds in 5ms
increments. This performs better in testing. I rarely saw a counter
less than 13 (2 failed retries). Current logging starts after 6
failures and counts down until success or failure.
| author | John Rouillard <rouilj@ieee.org> |
|---|---|
| date | Thu, 16 Dec 2021 20:02:00 -0500 |
| parents | 2120f77554d5 |
| children | e1dfd21f2252 |
line wrap: on
line source
#! /usr/bin/env python ''' Usage: %s <tracker home> <N> Load up the indicated tracker with N issues and N/100 users. ''' from __future__ import print_function import sys, os, random from roundup import instance # open the instance if len(sys.argv) < 2: print("Error: Not enough arguments") print(__doc__.strip()%(sys.argv[0])) sys.exit(1) tracker_home = sys.argv[1] N = int(sys.argv[2]) # open the tracker tracker = instance.open(tracker_home) db = tracker.open('admin') db.tx_Source = "cli" priorities = db.priority.list() statuses = db.status.list() resolved_id = db.status.lookup('resolved') statuses.remove(resolved_id) names = ['alpha', 'beta', 'gamma', 'delta', 'epsilon', 'zeta', 'eta', 'theta', 'iota', 'kappa', 'lambda', 'mu', 'nu', 'xi', 'omicron', 'pi', 'rho'] titles = '''Lorem ipsum dolor sit amet, consectetuer adipiscing elit. Duis nibh purus, bibendum sed, condimentum ut, bibendum ut, risus. Fusce pede enim, nonummy sit amet, dapibus a, blandit eget, metus. Nulla risus. Vivamus tincidunt. Donec consequat convallis quam. Sed convallis vehicula felis. Aliquam laoreet, dui quis pharetra vehicula, magna justo. Euismod felis, eu adipiscing eros metus id tortor. Suspendisse et turpis. Aenean non felis. Nam egestas eros. Integer tellus quam, mattis ac, vestibulum sed, egestas quis, mauris. Nulla tincidunt diam sit amet dui. Nam odio mauris, dignissim vitae, eleifend eu, consectetuer id, risus. Suspendisse potenti. Donec tincidunt. Vestibulum gravida. Fusce luctus, neque id mattis fringilla, purus pede sodales pede. Quis ultricies urna odio sed orci.'''.splitlines() try: try: db.user.lookup('alpha0') except: # add some users M = N//100 for i in range(M): print('\ruser', i, ' ', end=' ') sys.stdout.flush() if i//17 == 0: db.user.create(username=names[i%17]) else: db.user.create(username=names[i%17]+str(i//17)) # assignable user list users = db.user.list() users.remove(db.user.lookup('anonymous')) print() # now create the issues for i in range(N): print('\rissue', i, ' ', end=' ') sys.stdout.flush() # in practise, about 90% of issues are resolved if random.random() > .9: status = random.choice(statuses) else: status = resolved_id db.issue.create( title=random.choice(titles), priority=random.choice(priorities), status=status, assignedto=random.choice(users)) if not i%1000: db.commit() print() db.commit() finally: db.close() # vim: set filetype=python ts=4 sw=4 et si
